HLEDGER(1) hledger User Manuals HLEDGER(1) NAME hledger - a robust, friendly plain text accounting app (command line version). SYNOPSIS hledger or hledger COMMAND [OPTS] [ARGS] DESCRIPTION hledger is a robust, user-friendly, cross-platform set of programs for tracking money, time, or any other commodity, using double-entry ac- counting and a simple, editable file format. hledger is inspired by and largely compatible with ledger(1), and largely interconvertible with beancount(1). This manual is for hledger's command line interface, version 1.50. It also describes the common options, file formats and concepts used by all hledger programs. It might accidentally teach you some bookkeep- ing/accounting as well! You don't need to know everything in here to use hledger productively, but when you have a question about function- ality, this doc should answer it. It is detailed, so do skip ahead or skim when needed. You can read it on hledger.org, or as an info manual or man page on your system. You can also open a built-in copy, at a point of interest, by running hledger --man [CMD], hledger --info [CMD] or hledger help [TOPIC]. (And for shorter help, try hledger --tldr [CMD].) The main function of the hledger CLI is to read plain text files de- scribing financial transactions, crunch the numbers, and print a useful report on the terminal (or save it as HTML, CSV, JSON or SQL). Many reports are available, as subcommands. hledger will also detect other hledger-* executables as extra subcommands. hledger usually reads from (and appends to) a journal file specified by the LEDGER_FILE environment variable (defaulting to $HOME/.hledger.journal); or you can specify files with -f options. It can also read timeclock files, timedot files, or any CSV/SSV/TSV file with a date field. Here is a small journal file describing one transaction: 2015-10-16 bought food expenses:food $10 assets:cash Transactions are dated movements of money (etc.) between two or more accounts: bank accounts, your wallet, revenue/expense categories, peo- ple, etc. You can choose any account names you wish, using : to indi- cate subaccounts. There must be at least two spaces between account name and amount. Positive amounts are inflow to that account (debit), negatives are outflow from it (credit). (Some reports show revenue, liability and equity account balances as negative numbers as a result; this is normal.) hledger's add command can help you add transactions, or you can install other data entry UIs like hledger-web or hledger-iadd. For more exten- sive/efficient changes, use a text editor: Emacs + ledger-mode, VIM + vim-ledger, or VS Code + hledger-vscode are some good choices (see https://hledger.org/editors.html). To get started, run hledger add and follow the prompts, or save some entries like the above in $HOME/.hledger.journal, then try commands like: $ hledger print -x $ hledger aregister assets $ hledger balance $ hledger balancesheet $ hledger incomestatement Run hledger to list the commands. See also the "Starting a journal file" and "Setting opening balances" sections in PART 5: COMMON TASKS. PART 1: USER INTERFACE Input hledger reads one or more data files, each time you run it. You can specify a file with -f, like so $ hledger -f FILE [-f FILE2 ...] print Files are most often in hledger's journal format, with the .journal file extension (.hledger or .j also work); these files describe trans- actions, like an accounting general journal. When no file is specified, hledger looks for .hledger.journal in your home directory. But most people prefer to keep financial files in a dedicated folder, perhaps with version control. Also, starting a new journal file each year is common (it's not required, but helps keep things fast and or- ganised). So we usually configure a different journal file, by setting the LEDGER_FILE environment variable, to something like ~/fi- nance/2023.journal. For more about how to do that on your system, see Common tasks > Setting LEDGER_FILE. Text encoding hledger expects non-ascii input to be decodable with the system lo- cale's text encoding. (For CSV/SSV/TSV files, this can be overridden by the encoding CSV rule.) So, trying to read non-ascii files which have the wrong text encoding, or when no system locale is configured, will fail. To fix this, con- figure your system locale appropriately, and/or convert the files to your system's text encoding (using iconv on unix, or powershell or notepad on Windows). See Install: Text encoding for more tips. hledger's output will use the system locale's encoding. hledger's docs and example files mostly use UTF-8 encoding. Data formats Usually the data file is in hledger's journal format, but it can be in any of the supported file formats, which currently are: Reader: Reads: Automatically used for files with extensions: ----------------------------------------------------------------------------- journal hledger journal files and some .journal .j .hledger Ledger journals, for transactions .ledger timeclock timeclock files, for precise time .timeclock logging timedot timedot files, for approximate .timedot time logging csv Comma- or other delimiter-sepa- .csv rated values, for data import ssv Semicolon separated values .ssv tsv Tab separated values .tsv rules CSV/SSV/TSV/other separated val- .rules ues, alternate way These formats are described in more detail below. hledger detects the format automatically based on the file extensions shown above. If it can't recognise the file extension, it assumes journal format. So for non-journal files, it's important to use a recognised file extension, so as to either read successfully or to show relevant error messages. You can also force a specific reader/format by prefixing the file path with the format and a colon. Eg, to read a .dat file containing tab separated values: $ hledger -f tsv:/some/file.dat stats Standard input The file name - means standard input: $ cat FILE | hledger -f- print If reading non-journal data in this way, you'll need to write the for- mat as a prefix, like timeclock: here: $ echo 'i 2009/13/1 08:00:00' | hledger print -f timeclock:- Multiple files You can specify multiple -f options, to read multiple files as one big journal. When doing this, note that certain features (described below) will be affected: o Balance assertions will not see the effect of transactions in previ- ous files. (Usually this doesn't matter as each file will set the corresponding opening balances.) o Some directives will not affect previous or subsequent files. If needed, you can work around these by using a single parent file which includes the others, or concatenating the files into one, eg: cat a.journal b.journal | hledger -f- CMD. Strict mode hledger checks input files for valid data. By default, the most impor- tant errors are detected, while still accepting easy journal files without a lot of declarations: o Are the input files parseable, with valid syntax ? o Are all transactions balanced ? o Do all balance assertions pass ? With the -s/--strict flag, additional checks are performed: o Are all accounts posted to, declared with an account directive ? (Account error checking) o Are all commodities declared with a commodity directive ? (Commodity error checking) o Are all commodity conversions declared explicitly ? You can use the check command to run individual checks - the ones listed above and some more. Commands hledger provides various subcommands for getting things done. Most of these commands do not change the journal file; they just read it and output a report. A few commands assist with adding data and file man- agement. Some often-used commands are add, print, register, bal- ancesheet and incomestatement. To show a summary of commands, run hledger with no arguments. You can see the same commands summary at the start of PART 4: COMMANDS below. To use a particular command, run hledger CMD [CMDOPTS] [CMDARGS], o CMD is the full command name, or its standard abbreviation shown in the commands list, or any unambiguous prefix of the name. o CMDOPTS are command-specific options, if any. Command-specific op- tions must be written after the command name. Eg: hledger print -x. o CMDARGS are additional arguments to the command, if any. Most hledger commands accept arguments representing a query, to limit the data in some way. Eg: hledger reg assets:checking. To list a command's options, arguments, and documentation in the termi- nal, run hledger CMD -h. Eg: hledger bal -h. Add-on commands In addition to the built-in commands, you can install add-on commands, which will also appear in hledger's commands list. Some of these can be installed as separate packages; others can be found in hledger's bin/ directory, documented at https://hledger.org/scripts.html. Add-on commands are programs or scripts in your shell's PATH, whose name starts with "hledger-" and ends with no extension or a recognised extension (".bat", ".com", ".exe", ".hs", ".js", ".lhs", ".lua", ".php", ".pl", ".py", ".rb", ".rkt", or ".sh"), and (on unix and mac) which has executable permission for the current user. You can run add-on commands directly: hledger-ui --watch. Or you can run them with hledger, like built-in commands: hledger ui --watch. In this case hledger's config file will be used, so you can set custom options for the addon there. (Before hledger 1.50, an -- argument was needed before addon options, but not any more.) Options Run hledger -h to see general command line help. Options can be writ- ten either before or after the command name. These options are spe- cific to the hledger CLI: Flags: --conf=CONFFILE Use extra options defined in this config file. If not specified, searches upward and in XDG config dir for hledger.conf (or .hledger.conf in $HOME). -n --no-conf ignore any config file And the following general options are common to most hledger commands: General input/data transformation flags: -f --file=[FMT:]FILE Read data from FILE, or from stdin if FILE is -, inferring format from extension or a FMT: prefix. Can be specified more than once. If not specified, reads from $LEDGER_FILE or $HOME/.hledger.journal. --rules=RULESFILE Use rules defined in this rules file for converting subsequent CSV/SSV/TSV files. If not specified, uses FILE.csv.rules for each FILE.csv. --alias=A=B|/RGX/=RPL transform account names from A to B, or by replacing regular expression matches --auto generate extra postings by applying auto posting rules ("=") to all transactions --forecast[=PERIOD] Generate extra transactions from periodic rules ("~"), from after the latest ordinary transaction until 6 months from now. Or, during the specified PERIOD (the equals is required). Auto posting rules will also be applied to these transactions. In hledger-ui, also make future-dated transactions visible at startup. -I --ignore-assertions don't check balance assertions by default --txn-balancing=... how to check that transactions are balanced: 'old': use global display precision 'exact': use transaction precision (default) --infer-costs infer conversion equity postings from costs --infer-equity infer costs from conversion equity postings --infer-market-prices infer market prices from costs --pivot=TAGNAME use a different field or tag as account names -s --strict do extra error checks (and override -I) --verbose-tags add tags indicating generated/modified data General output/reporting flags (supported by some commands): -b --begin=DATE include postings/transactions on/after this date -e --end=DATE include postings/transactions before this date (with a report interval, will be adjusted to following subperiod end) -D --daily multiperiod report with 1 day interval -W --weekly multiperiod report with 1 week interval -M --monthly multiperiod report with 1 month interval -Q --quarterly multiperiod report with 1 quarter interval -Y --yearly multiperiod report with 1 year interval -p --period=PERIODEXP set begin date, end date, and/or report interval, with more flexibility --today=DATE override today's date (affects relative dates) --date2 match/use secondary dates instead (deprecated) -U --unmarked include only unmarked postings/transactions -P --pending include only pending postings/transactions -C --cleared include only cleared postings/transactions (-U/-P/-C can be combined) -R --real include only non-virtual postings -E --empty Show zero items, which are normally hidden. In hledger-ui & hledger-web, do the opposite. --depth=DEPTHEXP if a number (or -NUM): show only top NUM levels of accounts. If REGEXP=NUM, only apply limiting to accounts matching the regular expression. -B --cost show amounts converted to their cost/sale amount -V --market Show amounts converted to their value at period end(s) in their default valuation commodity. Equivalent to --value=end. -X --exchange=COMM Show amounts converted to their value at period end(s) in the specified commodity. Equivalent to --value=end,COMM. --value=WHEN[,COMM] show amounts converted to their value on the specified date(s) in their default valuation commodity or a specified commodity. WHEN can be: 'then': value on transaction dates 'end': value at period end(s) 'now': value today YYYY-MM-DD: value on given date -c --commodity-style=S Override a commodity's display style. Eg: -c '.' or -c '1.000,00 EUR' --pretty[=YN] Use box-drawing characters in text output? Can be 'y'/'yes' or 'n'/'no'. If YN is specified, the equals is required. General help flags: -h --help show command line help --tldr show command examples with tldr --info show the manual with info --man show the manual with man --version show version information --debug=[1-9] show this much debug output (default: 1) --pager=YN use a pager when needed ? y/yes (default) or n/no --color=YNA --colour use ANSI color ? y/yes, n/no, or auto (default) Usually hledger accepts any unambiguous flag prefix, eg you can write --tl instead of --tldr or --dry instead of --dry-run. You can combine short flags which don't take arguments, eg you can write -MAST instead of -M -A -S -T. Flags requiring an argument can't be combined in this way (-If FILE won't work). If the same option appears more than once in a command line, usually the last (right-most) wins. Similarly, if mutually exclusive flags are used together, the right-most wins. (When flags are mutually exclu- sive, they'll usually have a group prefix in --help.) With most commands, arguments are interpreted as a hledger query which filter the data. Some queries can be expressed either with options or with arguments. Below are more tips for using the command line interface - feel free to skip these until you need them. Special characters Here we touch on shell escaping/quoting rules, and give some examples. This is a slightly complicated topic which you may not need at first, but you should be aware of it, so you can return here when needed. If you are able to minimise the use of special characters in your data, you won't need escaping as much, and your command lines will be sim- pler. For example, avoiding spaces in account names, and using an ISO-4217 currency code like USD instead of the $ currency symbol, can be helpful. But if you want to use spaced account names and $, go right ahead; es- caping isn't a big deal. Escaping shell special characters At the command line, characters which have special meaning for your shell must be "shell-escaped" (AKA "quoted") if you want hledger to see them. Often these include space, <, >, (, ), |, \, $ and/or %. For example, to match an account name containing the phrase "credit card", don't write this: $ hledger register credit card In that command, "credit" and "card" are treated as separate query ar- guments (described below), so this would match accounts containing ei- ther word. Instead, enclose the phrase in double or single quotes: $ hledger register "credit card" In Unix shells, writing a backslash before the character can also work. Eg: $ hledger register credit\ card Some shell characters still have a special meaning inside double quotes, such as the dollar sign ($). Eg in "assets:$account", the bash shell would replace $account with the value of a shell variable with that name. When you don't want that, use single quotes, which escape more strongly: $ hledger balance 'assets:$account' Escaping on Windows If you are using hledger in a Powershell or Command window on Microsoft Windows, the escaping rules are different: o In a Powershell window (powershell, blue background), you must use double quotes or single quotes (not backslash). o In a Command window (cmd, black background), you must use double quotes (not single quotes or backslash). The next two sections were written for Unix-like shells, so might need to be adapted if you're using cmd or powershell. (Edits welcome.) Escaping regular expression special characters Many hledger arguments are regular expressions (described below), and these too have characters which cause special effects. Some of those characters are ., ^, $, [, ], (, ), |, and \. When you don't want these to cause special effects, you can "regex-escape" them by writing \ (a backslash) before them. But since backslash is also special to the shell, you may need to also shell-escape the backslashes. Eg, in the bash shell, to match a literal $ sign, you could write: $ hledger balance cur:\\$ or: $ hledger balance 'cur:\$' (The dollar sign is regex-escaped by the backslash preceding it. Then that backslash is shell-escaped by another backslash, or by single quotes.) Escaping add-on arguments When you run an external add-on command with hledger (described below), any options or arguments being passed through to the add-on executable lose one level of shell-escaping, so you must add an extra level of shell-escaping to compensate. Eg, in the bash shell, to run the ui add-on and match a literal $ sign, you need to write: $ hledger ui cur:'\\$' or: $ hledger ui cur:\\\\$ If you are wondering why four backslashes: o $ is unescaped o \$ is regex-escaped o \\$ is regex-escaped, then shell-escaped o \\\\$ is regex-escaped, then shell-escaped, then both slashes are shell-escaped once more for hledger argument pass-through. Or you can avoid such triple-escaping, by running the add-on executable directly: $ hledger-ui cur:\\$ Escaping in other situations hledger options and arguments are sometimes used in places other than the command line, with different escaping rules. For example, back- slash-quoting generally does not work there. Here are some more tips. In Windows cmd Use double quotes In Windows power- Use single or double quotes shell In hledger-ui's Use single or double quotes filter prompt In hledger-web's Use single or double quotes search form In an argument Don't use spaces, don't shell-escape, do regex-es- file cape when needed In a config file Use single or double quotes, and enclose the whole argument ("desc:a b" not desc:"a b") In ghci (the Use double quotes, and enclose the whole argument Haskell REPL) Using a wild card When escaping a special character is too much hassle (or impossible), you can often just write . (period) instead. In regular expressions, this means "accept any character here". Eg: $ hledger register credit.card Unicode characters hledger is expected to handle non-ascii characters correctly: o they should be parsed correctly in input files and on the command line, by all hledger tools (add, iadd, hledger-web's search/add/edit forms, etc.) o they should be displayed correctly by all hledger tools, and on-screen alignment should be preserved. This requires a well-configured environment. Here are some tips: o A system locale must be configured, which can decode the characters being used. This is essential - see Text encoding and Install: Text encoding. o Your terminal software (eg Terminal.app, iTerm, CMD.exe, xterm..) must support unicode. On Windows, you may need to use Windows Termi- nal. o The terminal must be using a font which includes the required unicode glyphs. o The terminal should be configured to display wide characters as dou- ble width (for report alignment). o On Windows, for best results you should run hledger in the same kind of environment in which it was built. Eg hledger built in the stan- dard CMD.EXE environment (like the binaries on our download page) might show display problems when run in a cygwin or msys terminal, and vice versa. (See eg #961). Regular expressions A regular expression (regexp) is a small piece of text where certain characters (like ., ^, $, +, *, (), |, [], \) have special meanings, forming a tiny language for matching text precisely - very useful in hledger and elsewhere. To learn all about them, visit regular-expres- sions.info. hledger supports regexps whenever you are entering a pattern to match something, eg in query arguments, account aliases, CSV if rules, hledger-web's search form, hledger-ui's / search, etc. You may need to wrap them in quotes, especially at the command line (see Special char- acters above). Here are some examples: Account name queries (quoted for command line use): Regular expression: Matches: ------------------- ------------------------------------------------------------ bank assets:bank, assets:bank:savings, expenses:art:banksy, ... :bank assets:bank:savings, expenses:art:banksy :bank: assets:bank:savings '^bank' none of those ( ^ matches beginning of text ) 'bank$' assets:bank ( $ matches end of text ) 'big \$ bank' big $ bank ( \ disables following character's special meaning ) '\bbank\b' assets:bank, assets:bank:savings ( \b matches word boundaries ) '(sav|check)ing' saving or checking ( (|) matches either alternative ) 'saving|checking' saving or checking ( outer parentheses are not needed ) 'savings?' saving or savings ( ? matches 0 or 1 of the preceding thing ) 'my +bank' my bank, my bank, ... ( + matches 1 or more of the preceding thing ) 'my *bank' mybank, my bank, my bank, ... ( * matches 0 or more of the preceding thing ) 'b.nk' bank, bonk, b nk, ... ( . matches any character ) Some other queries: desc:'amazon|amzn|audible' Amazon transactions cur:EUR amounts with commodity symbol containing EUR cur:'\$' amounts with commodity symbol containing $ cur:'^\$$' only $ amounts, not eg AU$ or CA$ cur:....? amounts with 4-or-more-character symbols tag:.=202[1-3] things with any tag whose value contains 2021, 2022 or 2023 Account name aliases: accept . instead of : as account separator: alias /\./=: replaces all periods in account names with colons Show multiple top-level accounts combined as one: --alias='/^[^:]+/=combined' ( [^:] matches any character other than : ) Show accounts with the second-level part removed: --alias '/^([^:]+):[^:]+/ = \1' match a top-level account and a second-level account and replace those with just the top-level account ( \1 in the replacement text means "whatever was matched by the first parenthesised part of the regexp" CSV rules: match CSV records containing dining-related MCC codes: if \?MCC581[124] Match CSV records with a specific amount around the end/start of month: if %amount \b3\.99 & %date (29|30|31|01|02|03)$ hledger's regular expressions hledger's regular expressions come from the regex-tdfa library. If they're not doing what you expect, it's important to know exactly what they support: 1. they are case insensitive 2. they are infix matching (they do not need to match the entire thing being matched) 3. they are POSIX ERE (extended regular expressions) 4. they also support GNU word boundaries (\b, \B, \<, \>) 5. backreferences are supported when doing text replacement in account aliases or CSV rules, where backreferences can be used in the re- placement string to reference capturing groups in the search regexp. Otherwise, if you write \1, it will match the digit 1. 6. they do not support mode modifiers ((?s)), character classes (\w, \d), or anything else not mentioned above. 7. they may not (I'm guessing not) properly support right-to-left or bidirectional text. Some things to note: o In the alias directive and --alias option, regular expressions must be enclosed in forward slashes (/REGEX/). Elsewhere in hledger, these are not required. o In queries, to match a regular expression metacharacter like $ as a literal character, prepend a backslash. Eg to search for amounts with the dollar sign in hledger-web, write cur:\$. o On the command line, some metacharacters like $ have a special mean- ing to the shell and so must be escaped at least once more. See Spe- cial characters. Argument files You can save a set of command line options and arguments in a file, and then reuse them by writing @FILENAME as a command line argument. Eg: hledger bal @foo.args. An argument file's format is more restrictive than the command line. Each line should contain just one option or argument. Don't use spaces except inside quotes; write = or nothing between a flag and its argu- ment. If you use quotes, they must enclose the whole line. For the special characters mentioned above, use one less level of quoting than you would at the command line. Config files With hledger 1.40+, you can save extra command line options and argu- ments in a more featureful hledger config file. Here's a small exam- ple: # General options are listed first, and used with hledger commands that support them. --pretty # Options following a `[COMMAND]` heading are used with that hledger command only. [print] --explicit --show-costs To use a config file, specify it with the --conf option. Its options will be inserted near the start of your command line, so you can over- ride them with command line options if needed. Or, you can set up an automatic config file that is used whenever you run hledger, by creating hledger.conf in the current directory or above, or .hledger.conf in your home directory (~/.hledger.conf), or hledger.conf in your XDG config directory (~/.con- fig/hledger/hledger.conf). Here is another example config you could start with: https://github.com/simonmichael/hledger/blob/master/hledger.conf.sample You can put not only options, but also arguments in a config file. If the first word in a config file's top (general) section does not begin with a dash (eg: print), it is treated as the command argument (over- riding any argument on the command line). On unix machines, you can add a shebang line at the top of a config file, set executable permission on the file, and use it like a script. Eg (the -S is needed on some operating systems): #!/usr/bin/env -S hledger --conf You can ignore config files by adding the -n/--no-conf flag to the com- mand line. This is useful when using hledger in scripts, or when trou- bleshooting. When both --conf and --no-conf options are used, the right-most wins. To inspect the processing of config files, use --debug or --debug=8. Or, run the setup command, which will display any active config files. (setup is not affected by config files itself, unlike other commands.) Warning! There aren't many hledger features that need a warning, but this is one! Automatic config files, while convenient, also make hledger less pre- dictable and dependable. It's easy to make a config file that changes a report's behaviour, or breaks your hledger-using scripts/applica- tions, in ways that will surprise you later. If you don't want this, 1. Just don't create a hledger.conf file on your machine. 2. Also be alert to downloaded directories which may contain a hledger.conf file. 3. Also if you are sharing scripts or examples or support, consider that others may have a hledger.conf file. Conversely, once you decide to use this feature, try to remember: 1. Whenever a hledger command does not work as expected, try it again with -n (--no-conf) to see if a config file was to blame. 2. Whenever you call hledger from a script, consider whether that call should use -n or not. 3. Be conservative about what you put in your config file; try to con- sider the effect on all your reports. 4. To troubleshoot the effect of config files, run with --debug or --debug 8. The config file feature was added in hledger 1.40. Shell completions If you use the bash or zsh shells, you can optionally set up con- text-sensitive autocompletion for hledger command lines. Try pressing hledger (should list all hledger commands) or hledger reg acct: (should list your top-level account names). If completions aren't working, or for more details, see Install > Shell completions. Output Output destination hledger commands send their output to the terminal by default. You can of course redirect this, eg into a file, using standard shell syntax: $ hledger print > foo.txt Some commands (print, register, stats, the balance commands) also pro- vide the -o/--output-file option, which does the same thing without needing the shell. Eg: $ hledger print -o foo.txt $ hledger print -o - # write to stdout (the default) Output format Some commands offer other kinds of output, not just text on the termi- nal. Here are those commands and the formats currently supported: command txt html csv/tsv fods beancount sql json -------------------------------------------------------------------------------------------- aregister Y Y Y Y Y balance Y Y Y Y Y balancesheet Y Y Y Y Y balancesheetequity Y Y Y Y Y cashflow Y Y Y Y Y incomestatement Y Y Y Y Y print Y Y Y Y Y Y Y register Y Y Y Y Y You can also see which output formats a command supports by running hledger CMD -h and looking for the -O/--output-format=FMT option, You can select the output format by using that option: $ hledger print -O csv # print CSV to standard output or by choosing a suitable filename extension with the -o/--out- put-file=FILE.FMT option: $ hledger balancesheet -o foo.csv # write CSV to foo.csv The -O option can be combined with -o to override the file extension if needed: $ hledger balancesheet -o foo.txt -O csv # write CSV to foo.txt Here are some notes about the various output formats. Text output This is the default: human readable, plain text report output, suitable for viewing with a monospace font in a terminal. If your data contains unicode or wide characters, you'll need a terminal and font that render those correctly. (This can be challenging on MS Windows.) Some reports (register, aregister) will normally use the full window width. If this isn't working or you want to override it, you can use the -w/--width option. Balance reports (balance, balancesheet, incomestatement...) use what- ever width they need. Multi-period multi-currency reports can often be wider than the window. Besides using a pager, helpful techniques for this situation include --layout=bare, -V, cur:, --transpose, --tree, --depth, --drop, switching to html output, etc. Box-drawing characters hledger draws simple table borders by default, to minimise the risk of display problems caused by a terminal/font not supporting box-drawing characters. But your terminal and font probably do support them, so we recommend using the --pretty flag to show prettier tables in the terminal. This is a good flag to add to your hledger config file. Colour hledger tries to automatically detect ANSI colour and text styling sup- port and use it when appropriate. (Currently, it is used rather mini- mally: some reports show negative numbers in red, and help output uses bold text for emphasis.) You can override this by setting the NO_COLOR environment variable to disable it, or by using the --color/--colour option, perhaps in your config file, with a y/yes or n/no value to force it on or off. Paging In unix-like environments, when displaying large output (in any output format) in the terminal, hledger tries to use a pager when appropriate. (You can disable this with the --pager=no option, perhaps in your con- fig file.) The pager shows one page of text at a time, and lets you scroll around to see more. While it is active, usually SPACE shows the next page, h shows help, and q quits. The home/end/page up/page down/cursor keys, and mouse scrolling, may also work. hledger will use the pager specified by the PAGER environment variable, otherwise less if available, otherwise more if available. (With one exception: hledger help -p TOPIC will always use less, so that it can scroll to the topic.) The pager is expected to display hledger's ANSI colour and text styling. If you see junk characters, you might need to configure your pager to handle ANSI codes. Or you could disable colour as described above. If you are using the less pager, hledger automatically appends a number of options to the LESS variable to enable ANSI colour and a number of other conveniences. (At the time of writing: --chop-long-lines --hilite-unread --ignore-case --no-init --quit-at-eof --quit-if-one-screen --RAW-CONTROL-CHARS --shift=8 --squeeze-blank-lines --use-backslash ). If these don't work well, you can set your preferred options in the HLEDGER_LESS variable, which will be used instead. HTML output HTML output can be styled by an optional hledger.css file in the same directory. HTML output will be a HTML fragment, not a complete HTML document. Like other hledger output, for non-ascii characters it will use the system locale's text encoding (see Text encoding). CSV / TSV output In CSV or TSV output, digit group marks (such as thousands separators) are disabled automatically. FODS output FODS is the OpenDocument Spreadsheet format as plain XML, as accepted by LibreOffice and OpenOffice. If you use their spreadsheet applica- tions, this is better than CSV because it works across locales (decimal point vs. decimal comma, character encoding stored in XML header, thus no problems with umlauts), it supports fixed header rows and columns, cell types (string vs. number vs. date), separation of number and currency (currency is displayed but the cell type is still a number ac- cessible for computation), styles (bold), borders. Btw. you can still extract CSV from FODS/ODS using various utilities like libreoffice --headless or ods2csv. Beancount output This is Beancount's journal format. You can use this to export your hledger data to Beancount, eg to use the Fava web app. hledger will try to adjust your data to suit Beancount, automatically. Be cautious and check the conversion until you are confident it is good. If you plan to export to Beancount often, you may want to follow its conventions, for a cleaner conversion: o use Beancount-friendly account names o use currency codes instead of currency symbols o use cost notation instead of equity conversion postings o avoid virtual postings, balance assignments, and secondary dates. There is one big adjustment you must handle yourself: for Beancount, the top level account names must be Assets, Liabilities, Equity, In- come, and/or Expenses. You can use account aliases to rewrite your ac- count names temporarily, if needed, as in this hledger2beancount.conf config file. 2024-12-20: Some more things not yet handled for you: o P directives are not converted automatically - convert those your- self. o Balance assignments are not converted (Beancount doesn't support them) - replace those with explicit amounts. Beancount account names Aside from the top-level names, hledger will adjust your account names to make valid Beancount account names, by capitalising each part, re- placing spaces with -, replacing other unsupported characters with C, prepending A to account name parts which don't begin with a letter or digit, and appending :A to account names which have only one part. Beancount commodity names hledger will adjust your commodity names to make valid Beancount com- modity/currency names, which must be 2-24 uppercase letters, digits, or ', ., _, -, beginning with a letter and ending with a letter or digit. hledger will convert known currency symbols to ISO 4217 currency codes, capitalise letters, replace spaces with -, replace other unsupported characters with C, and prepend or append C if needed. Beancount virtual postings Beancount doesn't allow virtual postings; if you have any, they will be omitted from beancount output. Beancount metadata hledger tags will be converted to Beancount metadata (except for tags whose name begins with _). Metadata names will be adjusted to be Bean- count-compatible: beginning with a lowercase letter, at least two char- acters long, and with unsupported characters encoded. Metadata values will use Beancount's string type. In hledger, objects can have the same tag repeated with multiple val- ues. Eg an assets:cash account might have both type:Asset and type:Cash tags. For Beancount these will be combined into one, with the values combined, comma separated. Eg: type: "Asset, Cash". Beancount costs Beancount doesn't allow redundant costs and conversion postings as hledger does. If you have any of these, the conversion postings will be omitted. Currently we support at most one cost + conversion post- ings group per transaction. Beancount operating currency Declaring an operating currency (or several) improves Beancount and Fava reports. Currently hledger will declare each currency used in cost amounts as an operating currency. If needed, replace these with your own declaration, like option "operating_currency" "USD" SQL output SQL output is expected to work at least with SQLite, MySQL and Post- gres. The SQL statements are expected to be executed in the empty database. If you already have tables created via SQL output of hledger, you would probably want to either clear data from these (via delete or truncate SQL statements) or drop the tables completely before import; otherwise your postings would be duplicated. For SQLite, it is more useful if you modify the generated id field to be a PRIMARY KEY. Eg: $ hledger print -O sql | sed 's/id serial/id INTEGER PRIMARY KEY AUTOINCREMENT NOT NULL/g' | ... This is not yet much used; feedback is welcome. JSON output Our JSON is rather large and verbose, since it is a faithful represen- tation of hledger's internal data types. To understand its structure, read the Haskell type definitions, which are mostly in https://github.com/simonmichael/hledger/blob/mas- ter/hledger-lib/Hledger/Data/Types.hs. hledger-web's OpenAPI specifi- cation may also be relevant. hledger stores numbers with sometimes up to 255 significant digits. This is too many digits for most JSON consumers, so in JSON output we round numbers to at most 10 decimal places. (We don't limit the number of integer digits.) If you find this causing problems, please let us know. Related: #1195 This is not yet much used; feedback is welcome. Commodity styles When displaying amounts, hledger infers a standard display style for each commodity/currency, as described below in Commodity display style. If needed, this can be overridden by a -c/--commodity-style option (ex- cept for cost amounts and amounts displayed by the print command, which are always displayed with all decimal digits). For example, the fol- lowing will force dollar amounts to be displayed as shown: $ hledger print -c '$1.000,0' This option can be repeated to set the display style for multiple com- modities/currencies. Its argument is as described in the commodity di- rective. In some cases hledger will adjust number formatting to improve their parseability (such as adding trailing decimal marks when needed). Debug output We intend hledger to be relatively easy to troubleshoot, introspect and develop. You can add --debug[=N] to any hledger command line to see additional debug output. N ranges from 1 (least output, the default) to 9 (maximum output). Typically you would start with 1 and increase until you are seeing enough. Debug output goes to stderr, and is not affected by -o/--output-file (unless you redirect stderr to stdout, eg: 2>&1). It will be interleaved with normal output, which can help re- veal when parts of the code are evaluated. To capture debug output in a log file instead, you can usually redirect stderr, eg: hledger bal --debug=3 2>hledger.log (This option doesn't work in a config file yet.) Environment These environment variables affect hledger: HLEDGER_LESS If less is your pager, this variable specifies the less options hledger should use. (Otherwise, LESS + custom options are used.) LEDGER_FILE The main journal file to use when not specified with -f/--file. Default: $HOME/.hledger.journal. NO_COLOR If this environment variable exists (with any value, including empty), hledger will not use ANSI color codes in terminal output, un- less overridden by an explicit --color=y or --colour=y option. PART 2: DATA FORMATS Journal hledger's usual data source is a plain text file containing journal en- tries in hledger journal format. If you're looking for a quick refer- ence, jump ahead to the journal cheatsheet (or use the table of con- tents at https://hledger.org/hledger.html). This file represents an accounting General Journal. The .journal file extension is most often used, though not strictly required. The jour- nal file contains a number of transaction entries, each describing a transfer of money (or any commodity) between two or more named ac- counts, in a simple format readable by both hledger and humans. hledger's journal format is compatible with most of Ledger's journal format, but not all of it. The differences and interoperation tips are described at hledger and Ledger. With some care, and by avoiding in- compatible features, you can keep your hledger journal readable by Ledger and vice versa. This can useful eg for comparing the behaviour of one app against the other. You can use hledger without learning any more about this file; just use the add or web or import commands to create and update it. Many users, though, edit the journal file with a text editor, and track changes with a version control system such as git. Editor add-ons such as ledger-mode or hledger-mode for Emacs, vim-ledger for Vim, and hledger-vscode for Visual Studio Code, make this easier, adding colour, formatting, tab completion, and useful commands. See Editor configura- tion at hledger.org for the full list. A hledger journal file can contain three kinds of thing: comment lines, transactions, and/or directives (including periodic transaction rules and auto posting rules). Understanding the journal file format will also give you a good understanding of hledger's data model. Here's a quick cheatsheet/overview, followed by detailed descriptions of each part. Journal cheatsheet # Here is the main syntax of hledger's journal format # (omitting extra Ledger compatibility syntax). ############################################################################### # 1. These are comment lines, for notes or temporarily disabling things. ; They begin with # or ; comment Or, lines can be enclosed within "comment" / "end comment". This is a block of commented lines. end comment # Some journal entries can have semicolon comments at end of line ; like this # Some of them require 2 or more spaces before the semicolon. ############################################################################### # 2. Directives customise processing or output in some way. # You don't need any directives to get started. # But they can add more error checking, or change how things are displayed. # They begin with a word, letter, or symbol. # They are most often placed at the top, before transactions. account assets ; Declare valid account names and display order. account assets:savings ; A subaccount. This one represents a bank account. account assets:checking ; Another. Note, 2+ spaces after the account name. account assets:receivable ; Accounting type is inferred from english names, account passifs ; or declared with a "type" tag, type:L account expenses ; type:X ; A follow-on comment line, indented. account expenses:rent ; Expense and revenue categories are also accounts. ; Subaccounts inherit their parent's type. commodity $0.00 ; Declare valid commodities and their display styles. commodity 1.000,00 EUR decimal-mark . ; The decimal mark used in this file (if ambiguous). payee Whole Foods ; Declare a valid payee name. tag trip ; Declare a valid tag name. P 2024-03-01 AAPL $179 ; Declare a market price for AAPL in $ on this date. include other.journal ; Include another journal file here. # Declare a recurring "periodic transaction", for budget/forecast reports ~ monthly set budget goals ; <- Note, 2+ spaces before the description. (expenses:rent) $1000 (expenses:food) $500 # Declare an auto posting rule, to modify existing transactions in reports = revenues:consulting liabilities:tax:2024:us *0.25 ; Add a tax liability & expense expenses:tax:2024:us *-0.25 ; for 25% of the revenue. ############################################################################### # 3. Transactions are what it's all about. # They are dated events, usually movements of money between 2 or more accounts. # They begin with a numeric date. # Here is their basic shape: # # DATE DESCRIPTION ; The transaction's date and optional description. # ACCOUNT1 AMOUNT ; A posting of an amount to/from this account, indented. # ACCOUNT2 AMOUNT ; A second posting, balancing the first. # ... ; More if needed. Amounts must sum to zero. # ; Note, 2+ spaces between account names and amounts. 2024-01-01 opening balances ; At the start, declare pre-existing balances this way. assets:savings $10000 ; Account names can be anything. lower case is easy to type. assets:checking $1000 ; assets, liabilities, equity, revenues, expenses are common. liabilities:credit card $-500 ; liabilities, equity, revenues balances are usually negative. equity:start ; One amount can be left blank. $-10500 is inferred here. ; Some of these accounts we didn't declare above, ; so -s/--strict would complain. 2024-01-03 ! (12345) pay rent ; Additional transaction comment lines, indented. ; There can be a ! or * after the date meaning "pending" or "cleared". ; There can be a parenthesised (code) after the date/status. ; Amounts' sign shows direction of flow. assets:checking $-500 ; Minus means removed from this account (credit). expenses:rent $500 ; Plus means added to this account (debit). ; Keeping transactions in date order is optional (but helps error checking). 2024-01-02 Gringott's Bank | withdrawal ; Description can be PAYEE | NOTE assets:bank:gold -10 gold assets:pouch 10 gold 2024-01-02 shopping expenses:clothing 1 gold expenses:wands 5 gold assets:pouch -6 gold 2024-01-02 receive gift revenues:gifts -3 "Chocolate Frogs" ; Complex commodity symbols assets:pouch 3 "Chocolate Frogs" ; must be in double quotes. 2024-01-15 buy some shares, in two lots ; Cost can be noted. assets:investments:2024-01-15 2.0 AAAA @ $1.50 ; @ means per-unit cost assets:investments:2024-01-15-02 3.0 AAAA @@ $4 ; @@ means total cost ; ^ Per-lot subaccounts are sometimes useful. assets:checking $-7 2024-01-15 assert some account balances on this date ; Balances can be asserted in any transaction, with =, for extra error checking. ; Assertion txns like this one can be made with hledger close --assert --show-costs ; assets:savings $0 = $10000 assets:checking $0 = $493 assets:bank:gold 0 gold = -10 gold assets:pouch 0 gold = 4 gold assets:pouch 0 "Chocolate Frogs" = 3 "Chocolate Frogs" assets:investments:2024-01-15 0.0 AAAA = 2.0 AAAA @ $1.50 assets:investments:2024-01-15-02 0.0 AAAA = 3.0 AAAA @@ $4 liabilities:credit card $0 = $-500 2024-02-01 note some event, or a transaction not yet fully entered, on this date ; Postings are not required. # Consistent YYYY-MM-DD date format is recommended, # but you can use . or / and omit leading zeros if you prefer. 2024.01.01 2024/1/1 Comments Lines in the journal will be ignored if they begin with a hash (#) or a semicolon (;). (See also Other syntax.) hledger will also ignore re- gions beginning with a comment line and ending with an end comment line (or file end). Here's a suggestion for choosing between them: o # for top-level notes o ; for commenting out things temporarily o comment for quickly commenting large regions (remember it's there, or you might get confused) Eg: # a comment line ; another commentline comment A multi-line comment block, continuing until "end comment" directive or the end of the current file. end comment Some hledger entries can have same-line comments attached to them, from ; (semicolon) to end of line. See Transaction comments, Posting com- ments, and Account comments below. Transactions Transactions are the main unit of information in a journal file. They represent events, typically a movement of some quantity of commodities between two or more named accounts. Each transaction is recorded as a journal entry, beginning with a sim- ple date in column 0. This can be followed by any of the following op- tional fields, separated by spaces: o a status character (empty, !, or *) o a code (any short number or text, enclosed in parentheses) o a description (any remaining text until end of line or a semicolon) o a comment (any remaining text following a semicolon until end of line, and any following indented lines beginning with a semicolon) o 0 or more indented posting lines, describing what was transferred and the accounts involved (indented comment lines are also allowed, but not blank lines or non-indented lines). Here's a simple journal file containing one transaction: 2008/01/01 income assets:bank:checking $1 income:salary $-1 Dates Simple dates Dates in the journal file use simple dates format: YYYY-MM-DD or YYYY/MM/DD or YYYY.MM.DD, with leading zeros optional. The year may be omitted, in which case it will be inferred from the context: the cur- rent transaction, the default year set with a Y directive, or the cur- rent date when the command is run. Some examples: 2010-01-31, 2010/01/31, 2010.1.31, 1/31. (The UI also accepts simple dates, as well as the more flexible smart dates documented in the hledger manual.) Posting dates You can give individual postings a different date from their parent transaction, by adding a posting comment containing a tag (see below) like date:DATE. This is probably the best way to control posting dates precisely. Eg in this example the expense should appear in May re- ports, and the deduction from checking should be reported on 6/1 for easy bank reconciliation: 2015/5/30 expenses:food $10 ; food purchased on saturday 5/30 assets:checking ; bank cleared it on monday, date:6/1 $ hledger -f t.j register food 2015-05-30 expenses:food $10 $10 $ hledger -f t.j register checking 2015-06-01 assets:checking $-10 $-10 DATE should be a simple date; if the year is not specified it will use the year of the transaction's date. The date: tag must have a valid simple date value if it is present, eg a date: tag with no value is not allowed. Status Transactions (or individual postings within a transaction) can have a status mark, which is a single character before the transaction de- scription (or posting account name), separated from it by a space, in- dicating one of three statuses: mark status ------------------ unmarked ! pending * cleared When reporting, you can filter by status with the -U/--unmarked, -P/--pending, and -C/--cleared flags (and you can combine these, eg -UP to match all except cleared things). Or you can use the status:, sta- tus:!, and status:* queries, or the U, P, C keys in hledger-ui. (Note: in Ledger the "unmarked" state is called "uncleared"; in hledger we renamed it to "unmarked" for semantic clarity.) Status marks are optional, but can be helpful eg for reconciling with real-world accounts. Some editor modes provide highlighting and short- cuts for working with status. Eg in Emacs ledger-mode, you can toggle transaction status with C-c C-e, or posting status with C-c C-c. What "uncleared", "pending", and "cleared" actually mean is up to you. Here's one suggestion: status meaning -------------------------------------------------------------------------- uncleared recorded but not yet reconciled; needs review pending tentatively reconciled (if needed, eg during a big reconcil- iation) cleared complete, reconciled as far as possible, and considered cor- rect With this scheme, you would use -PC to see the current balance at your bank, -U to see things which will probably hit your bank soon (like un- cashed checks), and no flags to see the most up-to-date state of your finances. Code After the status mark, but before the description, you can optionally write a transaction "code", enclosed in parentheses. This is a good place to record a check number, or some other important transaction id or reference number. Description After the date, status mark and/or code fields, the rest of the line (or until a comment is begun with ;) is the transaction's description. Here you can describe the transaction (called the "narration" in tradi- tional bookkeeping), or you can record a payee/payer name, or you can leave it empty. Transaction descriptions show up in print output and in register re- ports, and can be listed with the descriptions command. You can query by description with desc:DESCREGEX, or pivot on descrip- tion with --pivot desc. Payee and note Sometimes people want a dedicated payee/payer field that can be queried and checked more strictly. If you want that, you can write a | (pipe) character in the description. This divides it into a "payee" field on the left, and a "note" field on the right. (Either can be empty.) You can query these with payee:PAYEEREGEX and note:NOTEREGEX, list their values with the payees and notes commands, or pivot on payee or note. Note: in transactions with no | character, description, payee, and note all have the same value. Once a | is added, they become distinct. (If you'd like to change this behaviour, please propose it on the mail list.) If you want more strict error checking, you can declare the valid payee names with payee directives, and then enforce these with hledger check payees. (Note: because of the above, for this you'll need to ensure every transaction description contains a | and therefore a checkable payee name, even if it's empty.) Transaction comments Text following ;, after a transaction description, and/or on indented lines immediately below it, form comments for that transaction. They are reproduced by print but otherwise ignored, except they may contain tags, which are not ignored. 2012-01-01 something ; a transaction comment ; a second line of transaction comment expenses 1 assets Postings A posting is an addition of some amount to, or removal of some amount from, an account. Each posting line begins with at least one space or tab (2 or 4 spaces is common), followed by: o (optional) a status character (empty, !, or *), followed by a space o (required) an account name (any text, optionally containing single spaces, until end of line or a double space) o (optional) two or more spaces (or tabs) followed by an amount. If the amount is positive, it is being added to the account; if nega- tive, it is being removed from the account. The posting amounts in a transaction must sum up to zero, indicating that the inflows and outflows are equal. We call this a balanced transaction. (You can read more about the nitty-gritty details of "sum up to zero" in Transaction balancing below.) As a convenience, you can optionally leave one amount blank; hledger will infer what it should be so as to balance the transaction. Debits and credits The traditional accounting concepts of debit and credit of course exist in hledger, but we represent them with numeric sign, as described above. Positive and negative posting amounts represent debits and credits respectively. You don't need to remember that, but if you would like to - eg for helping newcomers or for talking with your accountant - here's a handy mnemonic: debit / plus / left / short words credit / minus / right / longer words The two space delimiter Be sure to notice the unusual separator between the account name and the following amount. Because hledger allows account names with spaces in them, you must separate the account name and amount (if any) by two or more spaces (or tabs). It's easy to forget at first. If you ever see the amount being treated as part of the account name, you'll know you probably need to add another space between them. Account names Accounts are the main way of categorising things in hledger. As in Double Entry Bookkeeping, they can represent real world accounts (such as a bank account), or more abstract categories such as "money borrowed from Frank" or "money spent on electricity". You can use any account names you like, but we usually start with the traditional accounting categories, which in english are assets, liabil- ities, equity, revenues, expenses. (You might see these referred to as A, L, E, R, X for short.) For more precise reporting, we usually divide the top level accounts into more detailed subaccounts, by writing a full colon between account name parts. For example, from the account names assets:bank:checking and expenses:food, hledger will infer this hierarchy of five accounts: assets assets:bank assets:bank:checking expenses expenses:food Shown as an outline, the hierarchical tree structure is more clear: assets bank checking expenses food hledger reports can summarise the account tree to any depth, so you can go as deep as you like with subcategories, but keeping your account names relatively simple may be best when starting out. Account names may be capitalised or not; they may contain letters, num- bers, symbols, or single spaces. Note, when an account name and an amount are written on the same line, they must be separated by two or more spaces (or tabs). Parentheses or brackets enclosing the full account name indicate vir- tual postings, described below. Parentheses or brackets internal to the account name have no special meaning. Account names can be altered temporarily or permanently by account aliases. Amounts After the account name, there is usually an amount. (Remember: between account name and amount, there must be two or more spaces.) hledger's amount format is flexible, supporting several international formats. Here are some examples. Amounts have a number (the "quan- tity"): 1 ..and usually a currency symbol or commodity name (more on this below), to the left or right of the quantity, with or without a separating space: $1 4000 AAPL 3 "green apples" Amounts can be preceded by a minus sign (or a plus sign, though plus is the default), The sign can be written before or after a left-side com- modity symbol: -$1 $-1 One or more spaces between the sign and the number are acceptable when parsing (but they won't be displayed in output): + $1 $- 1 Scientific E notation is allowed: 1E-6 EUR 1E3 Decimal marks A decimal mark can be written as a period or a comma: 1.23 1,23 Both of these are common in international number formats, so hledger is not biased towards one or the other. Because hledger also supports digit group marks (eg thousands separators), this means that a number like 1,000 or 1.000 containing just one period or comma is ambiguous. In such cases, hledger by default assumes it is a decimal mark, and will parse both of those as 1. To help hledger parse such ambiguous numbers more accurately, if you use digit group marks, we recommend declaring the decimal mark explic- itly. The best way is to add a decimal-mark directive at the top of each data file, like this: decimal-mark . Or you can declare it per commodity with commodity directives, de- scribed below. hledger also accepts numbers like 10. with no digits after the decimal mark (and will sometimes display numbers that way to disambiguate them - see Trailing decimal marks). Digit group marks In the integer part of the amount quantity (left of the decimal mark), groups of digits can optionally be separated by a digit group mark - a comma or period (whichever is not used as decimal mark), or a space (several Unicode space variants, like no-break space, are also ac- cepted). So these are all valid amounts in a journal file: $1,000,000.00 EUR 2.000.000,00 INR 9,99,99,999.00 1 000 000.00 ; <- ordinary space 1 000 000.00 ; <- no-break space Commodity Amounts in hledger have both a "quantity", which is a signed decimal number, and a "commodity", which is a currency symbol, stock ticker, or any word or phrase describing something you are tracking. If the commodity name contains non-letters (spaces, numbers, or punctu- ation), you must always write it inside double quotes ("green apples", "ABC123"). If you write just a bare number, that too will have a commodity, with name ""; we call that the "no-symbol commodity". Actually, hledger combines these single-commodity amounts into more powerful multi-commodity amounts, which are what it works with most of the time. A multi-commodity amount could be, eg: 1 USD, 2 EUR, 3.456 TSLA. In practice, you will only see multi-commodity amounts in hledger's output; you can't write them directly in the journal file. By default, the format of amounts in the journal influences how hledger displays them in output. This is explained in Commodity display style below. Costs After a posting amount, you can note its cost (when buying) or selling price (when selling) in another commodity, by writing either @ UNIT- PRICE or @@ TOTALPRICE after it. This indicates a conversion transac- tion, where one commodity is exchanged for another. (You might also see this called "transaction price" in hledger docs, discussions, or code; that term was directionally neutral and reminded that it is a price specific to a transaction, but we now just call it "cost", with the understanding that the transaction could be a purchase or a sale.) Costs are usually written explicitly with @ or @@, but can also be in- ferred automatically for simple multi-commodity transactions. Note, if costs are inferred, the order of postings is significant; the first posting will have a cost attached, in the commodity of the second. As an example, here are several ways to record purchases of a foreign currency in hledger, using the cost notation either explicitly or im- plicitly: 1. Write the price per unit, as @ UNITPRICE after the amount: 2009/1/1 assets:euros 100 @ $1.35 ; one hundred euros purchased at $1.35 each assets:dollars ; balancing amount is -$135.00 2. Write the total price, as @@ TOTALPRICE after the amount: 2009/1/1 assets:euros 100 @@ $135 ; one hundred euros purchased at $135 for the lot assets:dollars 3. Specify amounts for all postings, using exactly two commodities, and let hledger infer the price that balances the transaction. Note the effect of posting order: the price is added to first posting, making it 100 @@ $135, as in example 2: 2009/1/1 assets:euros 100 ; one hundred euros purchased assets:dollars $-135 ; for $135 Amounts can be converted to cost at report time using the -B/--cost flag; this is discussed more in the Cost reporting section. Note that the cost normally should be a positive amount, though it's not required to be. This can be a little confusing, see discussion at --infer-market-prices: market prices from transactions. Balance assertions hledger supports Ledger-style balance assertions in journal files. These look like, for example, = EXPECTEDBALANCE following a posting's amount. Eg here we assert the expected dollar balance in accounts a and b after each posting: 2013/1/1 a $1 = $1 b = $-1 2013/1/2 a $1 = $2 b $-1 = $-2 After reading a journal file, hledger will check all balance assertions and report an error if any of them fail. Balance assertions can pro- tect you from, eg, inadvertently disrupting reconciled balances while cleaning up old entries. You can disable them temporarily with the -I/--ignore-assertions flag, which can be useful for troubleshooting or for reading Ledger files. (Note: this flag currently does not disable balance assignments, described below). Assertions and ordering hledger calculates and checks an account's balance assertions in date order (and when there are multiple assertions on the same day, in parse order). Note this is different from Ledger, which checks assertions always in parse order, ignoring dates. This means in hledger you can freely reorder transactions, postings, or files, and balance assertions will usually keep working. The exception is when you reorder multiple postings on the same day, to the same ac- count, which have balance assertions; those will likely need updating. Assertions and multiple files If an account has transactions appearing in multiple files, balance as- sertions can still work - but only if those files are part of a hierar- chy made by include directives. If the same files are specified with two -f options on the command line, the assertions in the second will not see the balances from the first. To work around this, arrange your files in a hierarchy with include. Or, you could concatenate the files temporarily, and process them like one big file. Why does it work this way ? It might be related to hledger's goal of stable predictable reports. File hierarchy is considered "permanent", part of your data, while the order of command line options/arguments is not. We don't want transient changes to be able to change the meaning of the data. Eg it would be frustrating if tomorrow all your balance assertions broke because you wrote command line arguments in a differ- ent order. (Discussion welcome.) Assertions and costs Balance assertions ignore costs, and should normally be written without one: 2019/1/1 (a) $1 @ 1 = $1 We do allow costs to be written in balance assertion amounts, however, and print shows them, but they don't affect whether the assertion passes or fails. This is for backward compatibility (hledger's close command used to generate balance assertions with costs), and because balance assignments do use costs (see below). Assertions and commodities The balance assertions described so far are "single commodity balance assertions": they assert and check the balance in one commodity, ignor- ing any others that may be present. This is how balance assertions work in Ledger also. If an account contains multiple commodities, you can assert their bal- ances by writing multiple postings with balance assertions, one for each commodity: 2013/1/1 usd $-1 eur -1 both 2013/1/2 both 0 = $1 both 0 = 1 In hledger you can make a stronger "sole commodity balance assertion" by writing two equals signs (== EXPECTEDBALANCE). This also asserts that there are no other commodities in the account besides the asserted one (or at least, that their current balance is zero): 2013/1/1 usd $-1 == $-1 ; these sole commodity assertions succeed eur -1 == -1 both ;== $1 ; this one would fail because 'both' contains $ and It's less easy to make a "sole commodities balance assertion" (note the plural) - ie, asserting that an account contains two or more specified commodities and no others. It can be done by 1. isolating each commodity in a subaccount, and asserting those 2. and also asserting there are no commodities in the parent account itself: 2013/1/1 usd $-1 eur -1 both 0 == 0 ; nothing up my sleeve both:usd $1 == $1 ; a dollar here both:eur 1 == 1 ; a euro there Assertions and subaccounts All of the balance assertions above (both = and ==) are "subaccount-ex- clusive balance assertions"; they ignore any balances that exist in deeper subaccounts. In hledger you can make "subaccount-inclusive balance assertions" by adding a star after the equals (=* or ==*): 2019/1/1 equity:start assets:checking $10 assets:savings $10 assets $0 ==* $20 ; assets + subaccounts contains $20 and nothing else Assertions and status Balance assertions always consider postings of all statuses (unmarked, pending, or cleared); they are not affected by the -U/--unmarked / -P/--pending / -C/--cleared flags or the status: query. Assertions and virtual postings Balance assertions always consider both real and virtual postings; they are not affected by the --real/-R flag or real: query. Assertions and auto postings Balance assertions are affected by the --auto flag, which generates auto postings, which can alter account balances. Because auto postings are optional in hledger, accounts affected by them effectively have two balances. But balance assertions can only test one or the other of these. So to avoid making fragile assertions, either: o assert the balance calculated with --auto, and always use --auto with that file o or assert the balance calculated without --auto, and never use --auto with that file o or avoid balance assertions on accounts affected by auto postings (or avoid auto postings entirely). Assertions and precision Balance assertions compare the exactly calculated amounts, which are not always what is shown by reports. Eg a commodity directive may limit the display precision, but this will not affect balance asser- tions. Balance assertion failure messages show exact amounts. Assertions and hledger add Balance assertions can be included in the amounts given in add. All types of assertions are supported, and assertions can be used as in a normal journal file. All transactions, not just those that have an explicit assertion, are validated against the existing assertions in the journal. This means it is possible for an added transaction to fail even if its assertions are correct as of the transaction date. If this assertion checking is not desired, then it can be disabled with -I. However, balance assignments are currently not supported. Posting comments Text following ;, at the end of a posting line, and/or on indented lines immediately below it, form comments for that posting. They are reproduced by print but otherwise ignored, except they may contain tags, which are not ignored. 2012-01-01 expenses 1 ; a comment for posting 1 assets ; a comment for posting 2 ; a second comment line for posting 2 Transaction balancing How exactly does hledger decide when a transaction is balanced ? Espe- cially when it involves costs, which often are not exact, because of repeating decimals, or imperfect data from financial institutions ? In each commodity, hledger sums the transaction's posting amounts, after converting any with costs; then it checks if that sum is zero, when rounded to a suitable number of decimal digits - which we call the bal- ancing precision. Since version 1.50, hledger infers balancing precision in each transac- tion from the amounts in that transaction's journal entry (like Ledger). Ie, when checking the balance of commodity A, it uses the highest decimal precision seen for A in the journal entry (excluding cost amounts). This makes transaction balancing robust; any imbalances must be visibly accounted for in the journal entry, display precision can be freely increased with -c, and compatibility with Ledger and Beancount journals is good. Note that hledger versions before 1.50 worked differently: they allowed display precision to override the balancing precision. This masked small imbalances and caused fragility (see issue #2402). As a result, some journal entries (or CSV rules) that worked with hledger <1.50, are now rejected with an "unbalanced transaction" error. If you hit this problem, it's easy to fix: o You can restore the old behaviour, by adding --txn-balancing=old to the command or to your ~/.hledger.conf file. This lets you keep us- ing old journals unchanged, though without the above benefits. o Or you can fix the problem entries (recommended). There are three ways, use whichever seems best: 1. make cost amounts more precise (add more/better decimal digits) 2. or make non-cost amounts less precise (remove unnecessary decimal digits that are raising the precision) 3. or add a posting to absorb the imbalance (eg "expenses:rounding". Remember that one posting may omit the amount; that's convenient here.) Tags Tags are a way to add extra labels or data fields to transactions, postings, or accounts. They are usually a word or hyphenated word, im- mediately followed by a full colon, written within the comment of a transaction, a posting, or an account directive. (Yes, storing data in comments is slightly weird!) You can write each tag on its own comment line, or multiple tags on one line, separated by commas. Tags can also have a value, which is any text after the colon until the next comma or end of line, excluding surrounding whitespace. (hledger tag values can't contain commas.) If the same tag name appears multiple times in a comment, each name:value pair is preserved. An example: in this journal there are six tags, one of them with a value: account assets:checking ; accounttag: account expenses:food 2017/1/16 bought groceries ; transactiontag: ; transactiontag2: assets:checking $-1 ; posting-tag-1:, (belongs to the posting above) expenses:food $1 ; posting-tag-2:, posting-tag-3: with a value Querying with tags Tags are most often used to select a subset of data; you can match tagged things by tag name and or tag value with a tag: query. (See queries below.) When querying for tag names or values, note that postings inherit tags from their transaction and from their account, and transactions acquire tags from their postings. So in the example above, - the assets:check- ing posting effectively has four tags (one of its own, one from the ac- count, two from the transaction) - the expenses:food posting effec- tively has four tags (two of its own, two from the transaction) - the transaction effectively has all six tags (two of its own, and two from each posting) Displaying tags You can use the tags command to list tag names or values. The print command also shows tags. You can use --pivot to display tag values in other reports, in various ways (eg appended to account names, like pseudo subaccounts). When to use tags ? Tags provide more dimensions of categorisation, complementing accounts and transaction descriptions. When to use each of these is somewhat a matter of taste. Accounts have the most built-in support, and regex queries on descriptions are also quite powerful. So you may not need tags at all. But if you want to track multiple cross-cutting cate- gories, they can be a good fit. For example, you could tag trip-re- lated transactions with trip: YEAR:PLACE, without disturbing your usual account categories. Tag names What is allowed in a tag name ? Currently, most non-whitespace charac- ters. Eg : is a valid tag. For extra error checking, you can declare valid tag names with the tag directive, and then enforce these with the check command. But note that tags are detected quite loosely at present, sometimes where you didn't intend them. Eg ; see https://foo.com contains a https tag with value //foo.com. Special tags Some tag names have special significance to hledger. They are ex- plained elsewhere, but here's a quick reference: type -- declares an account's type date -- overrides a posting's date date2 -- overrides a posting's secondary date assert -- appears on txns generated by close --assert retain -- appears on txns generated by close --retain start -- appears on txns generated by close --migrate/--close/--open/--assign t -- appears on postings generated from timedot letters generated-transaction -- appears on txns generated by a periodic rule modified-transaction -- appears on txns which have had auto postings added generated-posting -- appears on generated postings cost-posting -- appears on postings which have (or could have) a cost, and which have equivalent conversion postings in the transaction conversion-posting -- appears on postings which are to a V/Conversion account and which have an equivalent cost posting in the transaction The second group above (generated-transaction, etc.) are normally hid- den, with a _ prefix added. This means print doesn't show them by de- fault; but you can still use them in queries. You can add the --ver- bose-tags flag to make them visible, which can be useful for trou- bleshooting. Directives Besides transactions, there is something else you can put in a journal file: directives. These are declarations, beginning with a keyword, that modify hledger's behaviour. Some directives can have more spe- cific subdirectives, indented below them. hledger's directives are similar to Ledger's in many cases, but there are also many differences. Directives are not required, but can be useful. Here are the main di- rectives: purpose directive -------------------------------------------------------------------------- READING DATA: Rewrite account names alias Comment out sections of the file comment Declare file's decimal mark, to help decimal-mark parse amounts accurately Include other data files include GENERATING DATA: Generate recurring transactions or bud- ~ get goals Generate extra postings on existing = transactions CHECKING FOR ERRORS: Define valid entities to provide more account, commodity, payee, tag error checking REPORTING: Declare accounts' type and display order account Declare commodity display styles commodity Declare market prices P Directives and multiple files Directives vary in their scope, ie which journal entries and which in- put files they affect. Most often, a directive will affect the follow- ing entries and included files if any, until the end of the current file - and no further. You might find this inconvenient! For example, alias directives do not affect parent or sibling files. But there are usually workarounds; for example, put alias directives in your top-most file, before including other files. The restriction, though it may be annoying at first, is in a good cause; it allows reports to be stable and deterministic, independent of the order of input. Without it, reports could show different numbers depending on the order of -f options, or the positions of include di- rectives in your files. Directive effects Here are all hledger's directives, with their effects and scope sum- marised - nine main directives, plus four others which we consider non-essential: di- what it does ends rec- at tive file end? -------------------------------------------------------------------------------------- ac- Declares an account, for checking all entries in all files; and N count its display order and type. Subdirectives: any text, ignored. alias Rewrites account names, in following entries until end of cur- Y rent file or end aliases. Command line equivalent: --alias com- Ignores part of the journal file, until end of current file or Y ment end comment. com- Declares up to four things: 1. a commodity symbol, for checking N,N,Y,Y mod- all amounts in all files 2. the display style for all amounts ity of this commodity 3. the decimal mark for parsing amounts of this commodity, in the rest of this file and its children, if there is no decimal-mark directive 4. the precision to use for balanced-transaction checking in this commodity, in this file and its children. Takes precedence over D. Subdirectives: format (ignored). Command line equivalent: -c/--commodity-style deci- Declares the decimal mark, for parsing amounts of all commodi- Y mal-mark ties in following entries until next decimal-mark or end of cur- rent file. Included files can override. Takes precedence over commodity and D. include Includes entries and directives from another file, as if they N were written inline. Command line alternative: multiple -f/--file payee Declares a payee name, for checking all entries in all files. N P Declares the market price of a commodity on some date, for value N reports. ~ Declares a periodic transaction rule that generates future N (tilde) transactions with --forecast and budget goals with balance --budget. Other syntax: apply Prepends a common parent account to all account names, in fol- Y account lowing entries until end of current file or end apply account. D Sets a default commodity to use for no-symbol amounts;and, if Y,Y,N,N there is no commodity directive for this commodity: its decimal mark, balancing precision, and display style, as above. Y Sets a default year to use for any yearless dates, in following Y entries until end of current file. = Declares an auto posting rule that generates extra postings on partly (equals) matched transactions with --auto, in current, parent, and child files (but not sibling files, see #1212). Other Other directives from Ledger's file format are accepted but ig- Ledger nored. direc- tives account directive account directives can be used to declare accounts (ie, the places that amounts are transferred from and to). Though not required, these dec- larations can provide several benefits: o They can document your intended chart of accounts, providing a refer- ence. o They can store additional account information as comments, or as tags which can be used to filter or pivot reports. o They can restrict which accounts may be posted to by transactions, eg in strict mode, which helps prevent errors. o They influence account display order in reports, allowing non-alpha- betic sorting (eg Revenues to appear above Expenses). o They can help hledger know your accounts' types (asset, liability, equity, revenue, expense), enabling reports like balancesheet and in- comestatement. o They help with account name completion (in hledger add, hledger-web, hledger-iadd, ledger-mode, etc.) They are written as the word account followed by a hledger-style ac- count name. Eg: account assets:bank:checking Ledger-style indented subdirectives are also accepted, but ignored: account assets:bank:checking format subdirective ; currently ignored Account comments Text following two or more spaces and ; at the end of an account direc- tive line, and/or following ; on indented lines immediately below it, form comments for that account. They are ignored except they may con- tain tags, which are not ignored. The two-space requirement for same-line account comments is because ; is allowed in account names. account assets:bank:checking ; same-line comment, at least 2 spaces before the semicolon ; next-line comment ; some tags - type:A, acctnum:12345 Account error checking By default, accounts need not be declared; they come into existence when a posting references them. This is convenient, but it means hledger can't warn you when you mis-spell an account name in the jour- nal. Usually you'll find that error later, as an extra account in bal- ance reports, or an incorrect balance when reconciling. In strict mode, enabled with the -s/--strict flag, or when you run hledger check accounts, hledger will report an error if any transaction uses an account name that has not been declared by an account direc- tive. Some notes: o The declaration is case-sensitive; transactions must use the correct account name capitalisation. o The account directive's scope is "whole file and below" (see direc- tives). This means it affects all of the current file, and any files it includes, but not parent or sibling files. The position of ac- count directives within the file does not matter, though it's usual to put them at the top. o Accounts can only be declared in journal files, but will affect in- cluded files of all types. o It's currently not possible to declare "all possible subaccounts" with a wildcard; every account posted to must be declared. o If you use the --infer-equity flag, you will also need declarations for the account names it generates. Account display order Account directives also cause hledger to display accounts in a particu- lar order, not just alphabetically. Eg, here is a conventional order- ing for the top-level accounts: account assets account liabilities account equity account revenues account expenses Now hledger displays them in that order: $ hledger accounts assets liabilities equity revenues expenses If there are undeclared accounts, those will be displayed last, in al- phabetical order. Sorting is done within each group of sibling accounts, at each level of the account tree. Eg, a declaration like account parent:child influ- ences child's position among its siblings. Note, it does not affect parent's position; for that, you need an ac- count parent declaration. Sibling accounts are always displayed together; hledger won't display x:y in between a:b and a:c. An account directive both declares an account as a valid posting tar- get, and declares its display order; you can't easily do one without the other. Account types hledger knows that in accounting there are three main account types: Asset A things you own Liability L things you owe Equity E owner's investment, balances the two above and two more representing changes in these: Revenue R inflows (also known as Income) Expense X outflows hledger also uses a couple of subtypes: Cash C liquid assets Conversion V commodity conver- sions equity As a convenience, hledger will detect these types automatically from english account names. But it's better to declare them explicitly by adding a type: tag in the account directives. The tag's value can be any of the types or one-letter abbreviations above. Here is a typical set of account type declarations. Subaccounts will inherit their parent's type, or can override it: account assets ; type: A account liabilities ; type: L account equity ; type: E account revenues ; type: R account expenses ; type: X account assets:bank ; type: C account assets:cash ; type: C account equity:conversion ; type: V This enables the easy balancesheet, balancesheetequity, cashflow and incomestatement reports, and querying by type:. Tips: o You can list accounts and their types, for troubleshooting: $ hledger accounts --types [ACCTPAT] [type:TYPECODES] [-DEPTH] [--positions] o It's a good idea to declare at least one account for each account type. Having some types declared and some inferred can disrupt cer- tain reports. o The rules for inferring types from account names are as follows (us- ing Regular expressions). If they don't work for you, just ignore them and declare your types with type: tags. If account's name contains this case insensitive regular expression | its type is --------------------------------------------------------------------|------------- ^assets?(:.+)?:(cash|bank|che(ck|que?)(ing)?|savings?|current)(:|$) | Cash ^assets?(:|$) | Asset ^(debts?|liabilit(y|ies))(:|$) | Liability ^equity:(trad(e|ing)|conversion)s?(:|$) | Conversion ^equity(:|$) | Equity ^(income|revenue)s?(:|$) | Revenue ^expenses?(:|$) | Expense o As mentioned above, subaccounts will inherit a type from their parent account. To be precise, an account's type is decided by the first of these that exists: 1. A type: declaration for this account. 2. A type: declaration in the parent accounts above it, preferring the nearest. 3. An account type inferred from this account's name. 4. An account type inferred from a parent account's name, preferring the nearest parent. 5. Otherwise, it will have no type. o Account aliases can disrupt account types. alias directive You can define account alias rules which rewrite your account names, or parts of them, before generating reports. This can be useful for: o expanding shorthand account names to their full form, allowing easier data entry and a less verbose journal o adapting old journals to your current chart of accounts o experimenting with new account organisations, like a new hierarchy o combining two accounts into one, eg to see their sum or difference on one line o customising reports Account aliases also rewrite account names in account directives. They do not affect account names being entered via hledger add or hledger-web. Account aliases are very powerful. They are generally easy to use cor- rectly, but you can also generate invalid account names with them; more on this below. See also Rewrite account names. Basic aliases To set an account alias, use the alias directive in your journal file. This affects all subsequent journal entries in the current file or its included files (but note: not sibling or parent files). The spaces around the = are optional: alias OLD = NEW Or, you can use the --alias 'OLD=NEW' option on the command line. This affects all entries. It's useful for trying out aliases interactively. OLD and NEW are case sensitive full account names. hledger will re- place any occurrence of the old account name with the new one. Subac- counts are also affected. Eg: alias checking = assets:bank:wells fargo:checking ; rewrites "checking" to "assets:bank:wells fargo:checking", or "checking:a" to "assets:bank:wells fargo:checking:a" Regex aliases There is also a more powerful variant that uses a regular expression, indicated by wrapping the pattern in forward slashes. (This is the only place where hledger requires forward slashes around a regular ex- pression.) Eg: alias /REGEX/ = REPLACEMENT or: $ hledger --alias '/REGEX/=REPLACEMENT' ... Any part of an account name matched by REGEX will be replaced by RE- PLACEMENT. REGEX is case-insensitive as usual. If you need to match a forward slash, escape it with a backslash, eg /\/=:. If REGEX contains parenthesised match groups, these can be referenced by the usual backslash and number in REPLACEMENT: alias /^(.+):bank:([^:]+):(.*)/ = \1:\2 \3 ; rewrites "assets:bank:wells fargo:checking" to "assets:wells fargo checking" REPLACEMENT continues to the end of line (or on command line, to end of option argument), so it can contain trailing whitespace. Combining aliases You can define as many aliases as you like, using journal directives and/or command line options. Recursive aliases - where an account name is rewritten by one alias, then by another alias, and so on - are allowed. Each alias sees the effect of previously applied aliases. In such cases it can be important to understand which aliases will be applied and in which order. For (each account name in) each journal entry, we apply: 1. alias directives preceding the journal entry, most recently parsed first (ie, reading upward from the journal entry, bottom to top) 2. --alias options, in the order they appeared on the command line (left to right). In other words, for (an account name in) a given journal entry: o the nearest alias declaration before/above the entry is applied first o the next alias before/above that will be be applied next, and so on o aliases defined after/below the entry do not affect it. This gives nearby aliases precedence over distant ones, and helps pro- vide semantic stability - aliases will keep working the same way inde- pendent of which files are being read and in which order. In case of trouble, adding --debug=6 to the command line will show which aliases are being applied when. Aliases and multiple files As explained at Directives and multiple files, alias directives do not affect parent or sibling files. Eg in this command, hledger -f a.aliases -f b.journal account aliases defined in a.aliases will not affect b.journal. In- cluding the aliases doesn't work either: include a.aliases 2023-01-01 ; not affected by a.aliases foo 1 bar This means that account aliases should usually be declared at the start of your top-most file, like this: alias foo=Foo alias bar=Bar 2023-01-01 ; affected by aliases above foo 1 bar include c.journal ; also affected end aliases directive You can clear (forget) all currently defined aliases (seen in the jour- nal so far, or defined on the command line) with this directive: end aliases Aliases can generate bad account names Be aware that account aliases can produce malformed account names, which could cause confusing reports or invalid print output. For exam- ple, you could erase all account names: 2021-01-01 a:aa 1 b $ hledger print --alias '/.*/=' 2021-01-01 1 The above print output is not a valid journal. Or you could insert an illegal double space, causing print output that would give a different journal when reparsed: 2021-01-01 old 1 other $ hledger print --alias old="new USD" | hledger -f- print 2021-01-01 new USD 1 other Aliases and account types If an account with a type declaration (see Declaring accounts > Account types) is renamed by an alias, normally the account type remains in ef- fect. However, renaming in a way that reshapes the account tree (eg renaming parent accounts but not their children, or vice versa) could prevent child accounts from inheriting the account type of their parents. Secondly, if an account's type is being inferred from its name, renam- ing it by an alias could prevent or alter that. If you are using account aliases and the type: query is not matching accounts as you expect, try troubleshooting with the accounts command, eg something like: $ hledger accounts --types -1 --alias assets=bassetts commodity directive The commodity directive performs several functions: 1. It declares which commodity symbols may be used in the journal, en- abling useful error checking with strict mode or the check command. See Commodity error checking below. 2. It declares how all amounts in this commodity should be displayed, eg how many decimals to show. See Commodity display style above. 3. (If no decimal-mark directive is in effect:) It sets the decimal mark to expect (period or comma) when parsing amounts in this com- modity, in this file and files it includes, from the directive until end of current file. See Decimal marks above. 4. It declares the precision with which this commodity's amounts should be compared when checking for balanced transactions, anywhere in this file and files it includes, until end of current file. Declaring commodities solves several common parsing/display problems, so we recommend it. Note that effects 3 and 4 above end at the end of the directive's file, and will not affect sibling or parent files. So if you are relying on them (especially 4) and using multiple files, placing your commodity directives in a top-level parent file might be important. Or, keep your decimal marks unambiguous and your entries well balanced and pre- cise. (Related: #793) Commodity directive syntax A commodity directive is normally the word commodity followed by a sam- ple amount (and optionally a comment). Only the amount's symbol and the number's format is significant. Eg: commodity $1000.00 commodity 1.000,00 EUR commodity 1 000 000.0000 ; the no-symbol commodity Commodities do not have tags (tags in the comment will be ignored). A commodity directive's sample amount must always include a period or comma decimal mark (this rule helps disambiguate decimal marks and digit group marks). If you don't want to show any decimal digits, write the decimal mark at the end: commodity 1000. AAAA ; show AAAA with no decimals Commodity symbols containing spaces, numbers, or punctuation must be enclosed in double quotes, as usual: commodity 1.0000 "AAAA 2023" Commodity directives normally include a sample amount, but can declare only a symbol (ie, just function 1 above): commodity $ commodity INR commodity "AAAA 2023" commodity "" ; the no-symbol commodity Commodity directives may also be written with an indented format subdi- rective, as in Ledger. The symbol is repeated and must be the same in both places. Other subdirectives are currently ignored: ; display indian rupees with currency name on the left, ; thousands, lakhs and crores comma-separated, ; period as decimal point, and two decimal places. commodity INR format INR 1,00,00,000.00 an unsupported subdirective ; ignored by hledger Commodity error checking In strict mode (-s/--strict) (or when you run hledger check commodi- ties), hledger will report an error if an undeclared commodity symbol is used. (With one exception: zero amounts are always allowed to have no commodity symbol.) It works like account error checking (described above). decimal-mark directive You can use a decimal-mark directive - usually one per file, at the top of the file - to declare which character represents a decimal mark when parsing amounts in this file. It can look like decimal-mark . or decimal-mark , This prevents any ambiguity when parsing numbers in the file, so we recommend it, especially if the file contains digit group marks (eg thousands separators). include directive You can pull in the content of additional files by writing an include directive, like this: include SOMEFILE This has the same effect as if SOMEFILE's content was inlined at this point. (With any include directives in SOMEFILE processed similarly, recursively.) Only journal files can include other files. They can include journal, timeclock or timedot files, but not CSV files. If the file path begins with a tilde, that means your home directory: include ~/main.journal. If it begins with a slash, it is an absolute path: include /home/user/main.journal. Otherwise it is relative to the including file's folder: include ../finances/main.journal. Also, the path may have a file type prefix to force a specific file format, overriding the file extension(s) (as described in Data for- mats): include timedot:notes/2023.md. The path may contain glob patterns to match multiple files. hledger's globs are similar to zsh's: ? to match any character; [a-z] to match any character in a range; * to match zero or more characters that aren't a path separator (like /); ** to match zero or more subdirecto- ries and/or zero or more characters at the start of a file name; etc. Also, hledger's globs always exclude the including file itself. So, you can do o include *.journal to include all other journal files in the current directory (excluding dot files) o include **.journal to include all other journal files in this direc- tory and below (excluding dot directories/files) o include timelogs/2???.timedot to include all timedot files named like a year number. There is a limitation: hledger's globs always exclude paths involving dot files or dot directories. This is a workaround for unavoidable dot directory traversal; you can disable it and revert to older behaviour with the --old-glob flag, for now. If you are using many, or deeply nested, include files, and have an er- ror that's hard to pinpoint: a good troubleshooting command is hledger files --debug=6 (or 7). P directive The P directive declares a market price, which is a conversion rate be- tween two commodities on a certain date. This allows value reports to convert amounts of one commodity to their value in another, on or after that date. These prices are often obtained from a stock exchange, cryptocurrency exchange, the or foreign exchange market. The format is: P DATE COMMODITY1SYMBOL COMMODITY2AMOUNT DATE is a simple date, COMMODITY1SYMBOL is the symbol of the commodity being priced, and COMMODITY2AMOUNT is the amount (symbol and quantity) of commodity 2 that one unit of commodity 1 is worth on this date. Ex- amples: # one euro was worth $1.35 from 2009-01-01 onward: P 2009-01-01 $1.35 # and $1.40 from 2010-01-01 onward: P 2010-01-01 $1.40 The -V, -X and --value flags use these market prices to show amount values in another commodity. See Value reporting. payee directive payee PAYEE NAME This directive can be used to declare a limited set of payees which may appear in transaction descriptions. The "payees" check will report an error if any transaction refers to a payee that has not been declared. Eg: payee Whole Foods ; a comment Payees do not have tags (tags in the comment will be ignored). To declare the empty payee name, use "". payee "" Ledger-style indented subdirectives, if any, are currently ignored. tag directive tag TAGNAME This directive can be used to declare a limited set of tag names al- lowed in tags. TAGNAME should be a valid tag name (no spaces). Eg: tag item-id Any indented subdirectives are currently ignored. The "tags" check will report an error if any undeclared tag name is used. It is quite easy to accidentally create a tag through normal use of colons in comments; if you want to prevent this, you can declare and check your tags . Periodic transactions The ~ directive declares a "periodic rule" which generates temporary extra transactions, usually recurring at some interval, when hledger is run with the --forecast flag. These "forecast transactions" are useful for forecasting future activity. They exist only for the duration of the report, and only when --forecast is used; they are not saved in the journal file by hledger. Periodic rules also have a second use: with the --budget flag they set budget goals for budgeting. Periodic rules can be a little tricky, so before you use them, read this whole section, or at least the following tips: 1. Two spaces accidentally added or omitted will cause you trouble - read about this below. 2. For troubleshooting, show the generated transactions with hledger print --forecast tag:generated or hledger register --forecast tag:generated. 3. Forecasted transactions will begin only after the last non-fore- casted transaction's date. 4. Forecasted transactions will end 6 months from today, by default. See below for the exact start/end rules. 5. period expressions can be tricky. Their documentation needs im- provement, but is worth studying. 6. Some period expressions with a repeating interval must begin on a natural boundary of that interval. Eg in weekly from DATE, DATE must be a monday. ~ weekly from 2019/10/1 (a tuesday) will give an error. 7. Other period expressions with an interval are automatically expanded to cover a whole number of that interval. (This is done to improve reports, but it also affects periodic transactions. Yes, it's a bit inconsistent with the above.) Eg: ~ every 10th day of month from 2023/01, which is equivalent to ~ every 10th day of month from 2023/01/01, will be adjusted to start on 2019/12/10. Periodic rule syntax A periodic transaction rule looks like a normal journal entry, with the date replaced by a tilde (~) followed by a period expression (mnemonic: ~ looks like a recurring sine wave.): # every first of month ~ monthly expenses:rent $2000 assets:bank:checking # every 15th of month in 2023's first quarter: ~ monthly from 2023-04-15 to 2023-06-16 expenses:utilities $400 assets:bank:checking The period expression is the same syntax used for specifying multi-pe- riod reports, just interpreted differently; there, it specifies report periods; here it specifies recurrence dates (the periods' start dates). Periodic rules and relative dates Partial or relative dates (like 12/31, 25, tomorrow, last week, next quarter) are usually not recommended in periodic rules, since the re- sults will change as time passes. If used, they will be interpreted relative to, in order of preference: 1. the first day of the default year specified by a recent Y directive 2. or the date specified with --today 3. or the date on which you are running the report. They will not be affected at all by report period or forecast period dates. Two spaces between period expression and description! If the period expression is followed by a transaction description, these must be separated by two or more spaces. This helps hledger know where the period expression ends, so that descriptions can not acciden- tally alter their meaning, as in this example: ; 2 or more spaces needed here, so the period is not understood as "every 2 months in 2023" ; || ; vv ~ every 2 months in 2023, we will review assets:bank:checking $1500 income:acme inc So, o Do write two spaces between your period expression and your transac- tion description, if any. o Don't accidentally write two spaces in the middle of your period ex- pression. Auto postings The = directive declares an "auto posting rule", which adds extra post- ings to existing transactions. (Remember, postings are the account name & amount lines below a transaction's date & description.) In the journal, an auto posting rule looks quite like a transaction, but instead of date and description it has = (mnemonic: "match") and a query, like this: = QUERY ACCOUNT AMOUNT ... Queries are just like command line queries; an account name substring is most common. Query terms containing spaces should be enclosed in single or double quotes. Each = rule works like this: when hledger is run with the --auto flag, wherever the QUERY matches a posting in the journal, the rule's post- ings are added to that transaction, immediately below the matched post- ing. Note these generated postings are temporary, existing only for the duration of the report, and only when --auto is used; they are not saved in the journal file by hledger. The postings can contain the special string %account which will be ex- panded to the account name of the matched account. Generated postings' amounts can depend on the matched posting's amount. So auto postings can be useful for, eg, adding tax postings with a standard percentage. AMOUNT can be: o a number with no commodity symbol, like 2. The matched posting's commodity symbol will be added to this. o a normal amount with a commodity symbol, like $2. This will be used as-is. o an asterisk followed by a number, like *2. This will multiply the matched posting's amount (and total price, if any) by the number. o an asterisk followed by an amount with commodity symbol, like *$2. This multiplies and also replaces the commodity symbol with this new one. Some examples: ; every time I buy food, schedule a dollar donation = expenses:food (liabilities:charity) $-1 ; when I buy a gift, also deduct that amount from a budget envelope subaccount = expenses:gifts assets:checking:gifts *-1 assets:checking *1 2017/12/1 expenses:food $10 assets:checking 2017/12/14 expenses:gifts $20 assets:checking $ hledger print --auto 2017-12-01 expenses:food $10 assets:checking (liabilities:charity) $-1 2017-12-14 expenses:gifts $20 assets:checking assets:checking:gifts -$20 assets:checking $20 Note that depending fully on generated data such as this has some draw- backs - it's less portable, less future-proof, less auditable by oth- ers, and less robust (eg your balance assertions will depend on whether you use or don't use --auto). An alternative is to use auto postings in "one time" fashion - use them to help build a complex journal entry, view it with hledger print --auto, and then copy that output into the journal file to make it permanent. Auto postings and multiple files An auto posting rule can affect any transaction in the current file, or in any parent file or child file. Note, currently it will not affect sibling files (when multiple -f/--file are used - see #1212). Auto postings and dates A posting date (or secondary date) in the matched posting, or (taking precedence) a posting date in the auto posting rule itself, will also be used in the generated posting. Auto postings and transaction balancing / inferred amounts / balance asser- tions Currently, auto postings are added: o after missing amounts are inferred, and transactions are checked for balancedness, o but before balance assertions are checked. Note this means that journal entries must be balanced both before and after auto postings are added. This changed in hledger 1.12+; see #893 for background. This also means that you cannot have more than one auto-posting with a missing amount applied to a given transaction, as it will be unable to infer amounts. Auto posting tags Automated postings will have some extra tags: o generated-posting:= QUERY - shows this was generated by an auto post- ing rule, and the query o _generated-posting:= QUERY - a hidden tag, which does not appear in hledger's output. This can be used to match postings generated "just now", rather than generated in the past and saved to the journal. Also, any transaction that has been changed by auto posting rules will have these tags added: o modified: - this transaction was modified o _modified: - a hidden tag not appearing in the comment; this transac- tion was modified "just now". Auto postings on forecast transactions only Tip: you can can make auto postings that will apply to forecast trans- actions but not recorded transactions, by adding tag:_generated-trans- action to their QUERY. This can be useful when generating new journal entries to be saved in the journal. Other syntax hledger journal format supports quite a few other features, mainly to make interoperating with or converting from Ledger easier. Note some of the features below are powerful and can be useful in special cases, but in general, features in this section are considered less important or even not recommended for most users. Downsides are mentioned to help you decide if you want to use them. Balance assignments Ledger-style balance assignments are also supported. These are like balance assertions, but with no posting amount on the left side of the equals sign; instead it is calculated automatically so as to satisfy the assertion. This can be a convenience during data entry, eg when setting opening balances: ; starting a new journal, set asset account balances 2016/1/1 opening balances assets:checking = $409.32 assets:savings = $735.24 assets:cash = $42 equity:opening balances or when adjusting a balance to reality: ; no cash left; update balance, record any untracked spending as a generic expense 2016/1/15 assets:cash = $0 expenses:misc The calculated amount depends on the account's balance in the commodity at that point (which depends on the previously-dated postings of the commodity to that account since the last balance assertion or assign- ment). Downsides: using balance assignments makes your journal less explicit; to know the exact amount posted, you have to run hledger or do the cal- culations yourself, instead of just reading it. Also balance assign- ments' forcing of balances can hide errors. These things make your fi- nancial data less portable, less future-proof, and less trustworthy in an audit. Balance assignments and costs A cost in a balance assignment will cause the calculated amount to have that cost attached: 2019/1/1 (a) = $1 @ 2 $ hledger print --explicit 2019-01-01 (a) $1 @ 2 = $1 @ 2 Balance assignments and multiple files Balance assignments handle multiple files like balance assertions. They see balance from other files previously included from the current file, but not from previous sibling or parent files. Bracketed posting dates For setting posting dates and secondary posting dates, Ledger's brack- eted date syntax is also supported: [DATE], [DATE=DATE2] or [=DATE2] in posting comments. hledger will attempt to parse any square-bracketed sequence of the 0123456789/-.= characters in this way. With this syn- tax, DATE infers its year from the transaction and DATE2 infers its year from DATE. Downsides: another syntax to learn, redundant with hledger's date:/date2: tags, and confusingly similar to Ledger's lot date syntax. D directive D AMOUNT This directive sets a default commodity, to be used for any subsequent commodityless amounts (ie, plain numbers) seen while parsing the jour- nal. This effect lasts until the next D directive, or the end of the current file. For compatibility/historical reasons, D also acts like a commodity di- rective (setting the commodity's decimal mark for parsing and display style for output). So its argument is not just a commodity symbol, but a full amount demonstrating the style. The amount must include a deci- mal mark (either period or comma). Eg: ; commodity-less amounts should be treated as dollars ; (and displayed with the dollar sign on the left, thousands separators and two decimal places) D $1,000.00 1/1 a 5 ; <- commodity-less amount, parsed as $5 and displayed as $5.00 b Interactions with other directives: For setting a commodity's display style, a commodity directive has highest priority, then a D directive. For detecting a commodity's decimal mark during parsing, decimal-mark has highest priority, then commodity, then D. For checking commodity symbols with the check command, a commodity di- rective is required (hledger check commodities ignores D directives). Downsides: omitting commodity symbols makes your financial data less explicit, less portable, and less trustworthy in an audit. It is usu- ally an unsustainable shortcut; sooner or later you will want to track multiple commodities. D is overloaded with functions redundant with commodity and decimal-mark. And it works differently from Ledger's D. apply account directive This directive sets a default parent account, which will be prepended to all accounts in following entries, until an end apply account direc- tive or end of current file. Eg: apply account home 2010/1/1 food $10 cash end apply account is equivalent to: 2010/01/01 home:food $10 home:cash $-10 account directives are also affected, and so is any included content. Account names entered via hledger add or hledger-web are not affected. Account aliases, if any, are applied after the parent account is prepended. Downsides: this can make your financial data less explicit, less portable, and less trustworthy in an audit. Y directive Y YEAR or (deprecated backward-compatible forms): year YEAR apply year YEAR The space is optional. This sets a default year to be used for subse- quent dates which don't specify a year. Eg: Y2009 ; set default year to 2009 12/15 ; equivalent to 2009/12/15 expenses 1 assets year 2010 ; change default year to 2010 2009/1/30 ; specifies the year, not affected expenses 1 assets 1/31 ; equivalent to 2010/1/31 expenses 1 assets Downsides: omitting the year (from primary transaction dates, at least) makes your financial data less explicit, less portable, and less trust- worthy in an audit. Such dates can get separated from their corre- sponding Y directive, eg when evaluating a region of the journal in your editor. A missing Y directive makes reports dependent on today's date. Secondary dates A secondary date is written after the primary date, following an equals sign: DATE1=DATE2. If the year is omitted, the primary date's year is assumed. When running reports, the primary (left side) date is used by default, but with the --date2 flag (--aux-date or--effective also work, for Ledger users), the secondary (right side) date will be used in- stead. The meaning of secondary dates is up to you. Eg it could be "primary is the bank's clearing date, secondary is the date the transaction was initiated, if different". In practice, this feature usually adds confusion: o You have to remember the primary and secondary dates' meaning, and follow that consistently. o It splits your bookkeeping into two modes, and you have to remember which mode is appropriate for a given report. o Usually your balance assertions will work with only one of these modes. o It makes your financial data more complicated, less portable, and less clear in an audit. o It interacts with every feature, creating an ongoing cost for imple- mentors. o It distracts new users and supporters. o Posting dates are simpler and work better. So secondary dates are officially deprecated in hledger, remaining only as a Ledger compatibility aid; we recommend using posting dates in- stead. Star comments Lines beginning with * (star/asterisk) are also comment lines. This feature allows Emacs users to insert org headings in their journal, al- lowing them to fold/unfold/navigate it like an outline when viewed with org mode. Downsides: another, unconventional comment syntax to learn. Decreases your journal's portability. And switching to Emacs org mode just for folding/unfolding meant losing the benefits of ledger mode; nowadays you can add outshine mode to ledger mode to get folding without losing ledger mode's features. Valuation expressions Ledger allows a valuation function or value to be written in double parentheses after an amount. hledger ignores these. Virtual postings A posting with parentheses around the account name, like (some:account) 10, is called an unbalanced virtual posting. These postings do not participate in transaction balancing. (And if you write them without an amount, a zero amount is always inferred.) These can occasionally be convenient for special circumstances, but they violate double entry bookkeeping and make your data less portable across applications, so many people avoid using them at all. A posting with brackets around the account name ([some:account]) is called a balanced virtual posting. The balanced virtual postings in a transaction must add up to zero, just like ordinary postings, but sepa- rately from them. These are not part of double entry bookkeeping ei- ther, but they are at least balanced. An example: 2022-01-01 buy food with cash, update budget envelope subaccounts, & something else assets:cash $-10 ; <- these balance each other expenses:food $7 ; <- expenses:food $3 ; <- [assets:checking:budget:food] $-10 ; <- and these balance each other [assets:checking:available] $10 ; <- (something:else) $5 ; <- this is not required to balance Ordinary postings, whose account names are neither parenthesised nor bracketed, are called real postings. You can exclude virtual postings from reports with the -R/--real flag or a real:1 query. Other Ledger directives These other Ledger directives are currently accepted but ignored. This allows hledger to read more Ledger files, but be aware that hledger's reports may differ from Ledger's if you use these. apply fixed COMM AMT apply tag TAG assert EXPR bucket / A ACCT capture ACCT REGEX check EXPR define VAR=EXPR end apply fixed end apply tag end apply year end tag eval / expr EXPR python PYTHONCODE tag NAME value EXPR --command-line-flags See also https://hledger.org/ledger.html for a detailed hledger/Ledger syntax comparison. Other cost/lot notations A slight digression for Ledger and Beancount users. Ledger has a number of cost/lot-related notations: o @ UNITCOST and @@ TOTALCOST o expresses a conversion rate, as in hledger o when buying, also creates a lot that can be selected at selling time o (@) UNITCOST and (@@) TOTALCOST (virtual cost) o like the above, but also means "this cost was exceptional, don't use it when inferring market prices". o {=UNITCOST} and {{=TOTALCOST}} (fixed price) o when buying, means "this cost is also the fixed value, don't let it fluctuate in value reports" o {UNITCOST} and {{TOTALCOST}} (lot price) o can be used identically to @ UNITCOST and @@ TOTALCOST, also cre- ates a lot o when selling, combined with @ ..., selects an existing lot by its cost basis. Does not check if that lot is present. o [YYYY/MM/DD] (lot date) o when buying, attaches this acquisition date to the lot o when selling, selects a lot by its acquisition date o (SOME TEXT) (lot note) o when buying, attaches this note to the lot o when selling, selects a lot by its note Currently, hledger o accepts any or all of the above in any order after the posting amount o supports @ and @@ o treats (@) and (@@) as synonyms for @ and @@ o and ignores the rest. (This can break transaction balancing.) Beancount has simpler notation and different behaviour: o @ UNITCOST and @@ TOTALCOST o expresses a cost without creating a lot, as in hledger o when buying (acquiring) or selling (disposing of) a lot, and com- bined with {...}: is not used except to document the cost/selling price o {UNITCOST} and {{TOTALCOST}} o when buying, expresses the cost for transaction balancing, and also creates a lot with this cost basis attached o when selling, o selects a lot by its cost basis o raises an error if that lot is not present or can not be selected unambiguously (depending on booking method configured) o expresses the selling price for transaction balancing o {}, {YYYY-MM-DD}, {"LABEL"}, {UNITCOST, "LABEL"}, {UNITCOST, YYYY-MM-DD, "LABEL"} o when selling, other combinations of date/cost/label, like the above, are accepted for selecting the lot. Currently, hledger o supports @ and @@ o accepts the {UNITCOST}/{{TOTALCOST}} notation, but ignores it o and rejects the rest. CSV hledger can read transactions from CSV (comma-separated values) files. More precisely, it can read DSV (delimiter-separated values), from a file or standard input. Comma-separated, semicolon-separated and tab-separated are the most common variants, and hledger will recognise these three automatically based on a .csv, .ssv or .tsv file name ex- tension or a csv:, ssv: or tsv: file path prefix. (To learn about producing CSV or TSV output, see Output format.) Each CSV file must be described by a corresponding rules file. This contains rules describing the CSV data (header line, fields layout, date format etc.), how to construct hledger transactions from it, and how to categorise transactions based on description or other attrib- utes. By default, hledger expects this rules file to be named like the CSV file, with an extra .rules extension added, in the same directory. Eg when asked to read foo/FILE.csv, hledger looks for foo/FILE.csv.rules. You can specify a different rules file with the --rules option. At minimum, the rules file must identify the date and amount fields, and often it also specifies the date format and how many header lines there are. Here's a simple CSV file and a rules file for it: Date, Description, Id, Amount 12/11/2019, Foo, 123, 10.23 # basic.csv.rules skip 1 fields date, description, , amount date-format %d/%m/%Y $ hledger print -f basic.csv 2019-11-12 Foo expenses:unknown 10.23 income:unknown -10.23 There's an introductory Tutorial: Import CSV data on hledger.org, and more CSV rules examples below, and a larger collection at https://github.com/simonmichael/hledger/tree/master/examples/csv. CSV rules cheatsheet The following kinds of rule can appear in the rules file, in any order. (Blank lines and lines beginning with # or ; or * are ignored.) source optionally declare which file to read data from archive optionally enable an archive of imported files encoding optionally declare which text encoding the data has separator declare the field separator, instead of rely- ing on file extension skip skip one or more header lines at start of file date-format declare how to parse CSV dates/date-times timezone declare the time zone of ambiguous CSV date-times newest-first improve txn order when: there are multiple records, newest first, all with the same date intra-day-reversed improve txn order when: same-day txns are in opposite order to the overall file decimal-mark declare the decimal mark used in CSV amounts, when ambiguous fields list name CSV fields for easy reference, and op- tionally assign their values to hledger fields Field assignment assign a CSV value or interpolated text value to a hledger field if block conditionally assign values to hledger fields, or skip a record or end (skip rest of file) if table conditionally assign values to hledger fields, using compact syntax balance-type select which type of balance assertions/as- signments to generate include inline another CSV rules file Working with CSV tips can be found below, including How CSV rules are evaluated. source If you tell hledger to read a csv file with -f foo.csv, it will look for rules in foo.csv.rules. Or, you can tell it to read the rules file, with -f foo.csv.rules, and it will look for data in foo.csv (since 1.30). These are mostly equivalent, but the second method pro- vides some extra features. For one, the data file can be missing, without causing an error; it is just considered empty. For more flexibility, add a source rule, which lets you specify a dif- ferent data file: source ./Checking1.csv If the file does not exist, it is just considered empty, without rais- ing an error. If you specify just a file name with no path, hledger will look for it in the ~/Downloads folder: source Checking1.csv You can use a glob pattern, to avoid specifying the file name exactly: source Checking1*.csv This has another benefit: if the pattern matches multiple files, hledger will read the newest (most recently modified) one. This avoids problems if you have downloaded a file multiple times without cleaning up. All this enables a convenient workflow where can you just download CSV files, then run hledger import rules/*. See also "Working with CSV > Reading files specified by rule". Data cleaning / generating commands After source's file pattern, you can write | (pipe) and a data cleaning command. If hledger's CSV rules aren't enough, you can pre-process the downloaded data here with a shell command or script, to make it more suitable for conversion. The command will be executed by your default shell, in the directory of the rules file, will receive the data file's content as standard input, and should output zero or more lines of character-separated-values, suitable for conversion by the CSV rules. Examples: source ./paypal.json | paypalcsv source data/simplefin.json | simplefincsv - 'chase.*card' source OfxDownload*.csv | grep -vE '^(([^,]*,){6}[^,]*|)$' | sort -t, -n +2 source History_for_Account_Z20144832*.csv # | grep -E '^([^,]*,){12}[^,]*$' | sed -E -e 's/^ //' -e 's/\.([0-9]),/.\10,/g' -e 's/,([0-9]+),/,\1.00,/g' Or, after source you can write | and a data generating command (with no file pattern before the |). This command receives no input, and should output zero or more lines of character-separated values, suitable for conversion by the CSV rules. Examples: source | paypaljson | paypalcsv source | paypalcsv data/paypal.json source | simplefinjson >data/simplefin.json && simplefincsv data/simplefin.json 'chase.*card' source | simplefincsv data/simplefin.json 'unify.*checking' (paypal* and simplefin* scripts are in bin/) Whenever hledger runs one of these commands, it will print the command on stderr. If the command produces error output, but exits success- fully, hledger will show the error output as a warning. If the command fails, hledger will fail and show the error output in the error mes- sage. Added in 1.50; experimental. archive With archive added to a rules file, the import command will archive each successfully processed data file or data command output in a nearby data/ directory. The archive file name will be based on the rules file and the data file's modification date and extension (or for a data-generating command, the current date and the ".csv" extension). The original data file, if any, will be removed. Also, in this mode import will prefer the oldest file matched by the source rule's glob pattern, not the newest. (So if there are multiple downloads, they will be imported and archived oldest first.) Archiving is optional, but it can be useful for troubleshooting your CSV rules, regenerating entries with improved rules, checking for vari- ations in your bank's CSV, etc. Added in 1.50; experimental. encoding encoding ENCODING hledger normally expects non-ascii text to be using the system locale's text encoding. If you need to read CSV files which have some other en- coding, you can do it by adding encoding ENCODING to your CSV rules. Eg: encoding iso-8859-1. The following encodings are supported: ascii, utf-8, utf-16, utf-32, iso-8859-1, iso-8859-2, iso-8859-3, iso-8859-4, iso-8859-5, iso-8859-6, iso-8859-7, iso-8859-8, iso-8859-9, iso-8859-10, iso-8859-11, iso-8859-13, iso-8859-14, iso-8859-15, iso-8859-16, cp1250, cp1251, cp1252, cp1253, cp1254, cp1255, cp1256, cp1257, cp1258, koi8-r, koi8-u, gb18030, macintosh, jis-x-0201, jis-x-0208, iso-2022-jp, shift-jis, cp437, cp737, cp775, cp850, cp852, cp855, cp857, cp860, cp861, cp862, cp863, cp864, cp865, cp866, cp869, cp874, cp932. Added in 1.42. separator You can use the separator rule to read other kinds of character-sepa- rated data. The argument is any single separator character, or the words tab or space (case insensitive). Eg, for comma-separated values (CSV): separator , or for semicolon-separated values (SSV): separator ; or for tab-separated values (TSV): separator TAB If the input file has a .csv, .ssv or .tsv file extension (or a csv:, ssv:, tsv: prefix), the appropriate separator will be inferred automat- ically, and you won't need this rule. skip skip N The word skip followed by a number (or no number, meaning 1) tells hledger to ignore this many non-empty lines at the start of the input data. You'll need this whenever your CSV data contains header lines. Note, empty and blank lines are skipped automatically, so you don't need to count those. skip has a second meaning: it can be used inside if blocks (described below), to skip one or more records whenever the condition is true. Records skipped in this way are ignored, except they are still required to be valid CSV. date-format date-format DATEFMT This is a helper for the date (and date2) fields. If your CSV dates are not formatted like YYYY-MM-DD, YYYY/MM/DD or YYYY.MM.DD, you'll need to add a date-format rule describing them with a strptime-style date parsing pattern - see https://hackage.haskell.org/pack- age/time/docs/Data-Time-Format.html#v:formatTime. The pattern must parse the CSV date value completely. Some examples: # MM/DD/YY date-format %m/%d/%y # D/M/YYYY # The - makes leading zeros optional. date-format %-d/%-m/%Y # YYYY-Mmm-DD date-format %Y-%h-%d # M/D/YYYY HH:MM AM some other junk # Note the time and junk must be fully parsed, though only the date is used. date-format %-m/%-d/%Y %l:%M %p some other junk Note currently there is no locale awareness for things like %b, and setting LC_TIME won't help. timezone timezone TIMEZONE When CSV contains date-times that are implicitly in some time zone other than yours, but containing no explicit time zone information, you can use this rule to declare the CSV's native time zone, which helps prevent off-by-one dates. When the CSV date-times do contain time zone information, you don't need this rule; instead, use %Z in date-format (or %z, %EZ, %Ez; see the formatTime link above). In either of these cases, hledger will do a time-zone-aware conversion, localising the CSV date-times to your current system time zone. If you prefer to localise to some other time zone, eg for reproducibility, you can (on unix at least) set the output timezone with the TZ environment variable, eg: $ TZ=-1000 hledger print -f foo.csv # or TZ=-1000 hledger import foo.csv timezone currently does not understand timezone names, except "UTC", "GMT", "EST", "EDT", "CST", "CDT", "MST", "MDT", "PST", or "PDT". For others, use numeric format: +HHMM or -HHMM. newest-first hledger tries to ensure that the generated transactions will be ordered chronologically, including same-day transactions. Usually it can auto-detect how the CSV records are ordered. But if it encounters CSV where all records are on the same date, it assumes that the records are oldest first. If in fact the CSV's records are normally newest first, like: 2022-10-01, txn 3... 2022-10-01, txn 2... 2022-10-01, txn 1... you can add the newest-first rule to help hledger generate the transac- tions in correct order. # same-day CSV records are newest first newest-first intra-day-reversed If CSV records within a single day are ordered opposite to the overall record order, you can add the intra-day-reversed rule to improve the order of journal entries. Eg, here the overall record order is newest first, but same-day records are oldest first: 2022-10-02, txn 3... 2022-10-02, txn 4... 2022-10-01, txn 1... 2022-10-01, txn 2... # transactions within each day are reversed with respect to the overall date order intra-day-reversed decimal-mark decimal-mark . or: decimal-mark , hledger automatically accepts either period or comma as a decimal mark when parsing numbers (cf Amounts). However if any numbers in the CSV contain digit group marks, such as thousand-separating commas, you should declare the decimal mark explicitly with this rule, to avoid misparsed numbers. fields list fields FIELDNAME1, FIELDNAME2, ... A fields list (the word fields followed by comma-separated field names) is optional, but convenient. It does two things: 1. It names the CSV field in each column. This can be convenient if you are referencing them in other rules, so you can say %SomeField instead of remembering %13. 2. Whenever you use one of the special hledger field names (described below), it assigns the CSV value in this position to that hledger field. This is the quickest way to populate hledger's fields and build a transaction. Here's an example that says "use the 1st, 2nd and 4th fields as the transaction's date, description and amount; name the last two fields for later reference; and ignore the others": fields date, description, , amount, , , somefield, anotherfield In a fields list, the separator is always comma; it is unrelated to the CSV file's separator. Also: o There must be least two items in the list (at least one comma). o Field names may not contain spaces. Spaces before/after field names are optional. o Field names may contain _ (underscore) or - (hyphen). o Fields you don't care about can be given a dummy name or an empty name. If the CSV contains column headings, it's convenient to use these for your field names, suitably modified (eg lower-cased with spaces re- placed by underscores). Sometimes you may want to alter a CSV field name to avoid assigning to a hledger field with the same name. Eg you could call the CSV's "bal- ance" field balance_ to avoid directly setting hledger's balance field (and generating a balance assertion). Field assignment HLEDGERFIELD FIELDVALUE Field assignments are the more flexible way to assign CSV values to hledger fields. They can be used instead of or in addition to a fields list (see above). To assign a value to a hledger field, write the field name (any of the standard hledger field/pseudo-field names, defined below), a space, followed by a text value on the same line. This text value may inter- polate CSV fields, referenced either by their 1-based position in the CSV record (%N) or by the name they were given in the fields list (%CSVFIELD), and regular expression match groups (\N). Some examples: # set the amount to the 4th CSV field, with " USD" appended amount %4 USD # combine three fields to make a comment, containing note: and date: tags comment note: %somefield - %anotherfield, date: %1 Tips: o Interpolation strips outer whitespace (so a CSV value like " 1 " be- comes 1 when interpolated) (#1051). o Interpolations always refer to a CSV field - you can't interpolate a hledger field. (See Referencing other fields below). Field names Note the two kinds of field names mentioned here, and used only in hledger CSV rules files: 1. CSV field names (CSVFIELD in these docs): you can optionally name the CSV columns for easy reference (since hledger doesn't yet auto- matically recognise column headings in a CSV file), by writing arbi- trary names in a fields list, eg: fields When, What, Some_Id, Net, Total, Foo, Bar 2. Special hledger field names (HLEDGERFIELD in these docs): you must set at least some of these to generate the hledger transaction from a CSV record, by writing them as the left hand side of a field as- signment, eg: date %When code %Some_Id description %What comment %Foo %Bar amount1 $ %Total or directly in a fields list: fields date, description, code, , amount1, Foo, Bar currency $ comment %Foo %Bar Here are all the special hledger field names available, and what hap- pens when you assign values to them: date field Assigning to date sets the transaction date. date2 field date2 sets the transaction's secondary date, if any. status field status sets the transaction's status, if any. code field code sets the transaction's code, if any. description field description sets the transaction's description, if any. comment field comment sets the transaction's comment, if any. commentN, where N is a number, sets the Nth posting's comment. You can assign multi-line comments by writing literal \n in the code. A comment starting with \n will begin on a new line. Comments can contain tags, as usual. Posting comments can also contain a posting date. A secondary date, or a year-less date, will be ignored. account field Assigning to accountN, where N is 1 to 99, sets the account name of the Nth posting, and causes that posting to be generated. Most often there are two postings, so you'll want to set account1 and account2. Typically account1 is associated with the CSV file, and is set once with a top-level assignment, while account2 is set based on each transaction's description, in conditional rules. If a posting's account name is left unset but its amount is set (see below), a default account name will be chosen (like "expenses:unknown" or "income:unknown"). amount field There are several ways to set posting amounts from CSV, useful in dif- ferent situations. 1. amount is the oldest and simplest. Assigning to this sets the amount of the first and second postings. In the second posting, the amount will be negated; also, if it has a cost attached, it will be converted to cost. 2. amount-in and amount-out work exactly like the above, but should be used when the CSV has two amount fields (such as "Debit" and "Credit", or "Inflow" and "Outflow"). Whichever field has a non-zero value will be used as the amount of the first and second postings. Here are some tips to avoid confusion: o It's not "amount-in for posting 1 and amount-out for posting 2", it is "extract a single amount from the amount-in or amount-out field, and use that for posting 1 and (negated) for posting 2". o Don't use both amount and amount-in/amount-out in the same rules file; choose based on whether the amount is in a single CSV field or spread across two fields. o In each record, at most one of the two CSV fields should contain a non-zero amount; the other field must contain a zero or noth- ing. o hledger assumes both CSV fields contain unsigned numbers, and it automatically negates the amount-out values. o If the data doesn't fit these requirements, you'll probably need an if rule (see below). 3. amountN (where N is a number from 1 to 99) sets the amount of only a single posting: the Nth posting in the transaction. You'll usually need at least two such assignments to make a balanced transaction. You can also generate more than two postings, to represent more com- plex transactions. The posting numbers don't have to be consecu- tive; with if rules, higher posting numbers can be useful to ensure a certain order of postings. 4. amountN-in and amountN-out work exactly like the above, but should be used when the CSV has two amount fields. This is analogous to amount-in and amount-out, and those tips also apply here. 5. Remember that a fields list can also do assignments. So in a fields list if you name a CSV field "amount", that counts as assigning to amount. (If you don't want that, call it something else in the fields list, like "amount_".) 6. The above don't handle every situation; if you need more flexibil- ity, use an if rule to set amounts conditionally. See "Working with CSV > Setting amounts" below for more on this and on amount-setting generally. currency field currency sets a currency symbol, to be prepended to all postings' amounts. You can use this if the CSV amounts do not have a currency symbol, eg if it is in a separate column. currencyN prepends a currency symbol to just the Nth posting's amount. balance field balanceN sets a balance assertion amount (or if the posting amount is left empty, a balance assignment) on posting N. balance is a compatibility spelling for hledger <1.17; it is equivalent to balance1. You can adjust the type of assertion/assignment with the balance-type rule (see below). See the Working with CSV tips below for more about setting amounts and currency. if block Rules can be applied conditionally, depending on patterns in the CSV data. This allows flexibility; in particular, it is how you can cate- gorise transactions, selecting an appropriate account name based on their description (for example). There are two ways to write condi- tional rules: "if blocks", described here, and "if tables", described below. An if block is the word if and one or more "matcher" expressions (can be a word or phrase), one per line, starting either on the same or next line; followed by one or more indented rules. Eg, if MATCHER RULE or if MATCHER MATCHER MATCHER RULE RULE If any of the matchers succeeds, all of the indented rules will be ap- plied. They are usually field assignments, but the following special rules may also be used within an if block: o skip - skips the matched CSV record (generating no transaction from it) o end - skips the rest of the current CSV file. Some examples: # if the record contains "groceries", set account2 to "expenses:groceries" if groceries account2 expenses:groceries # if the record contains any of these phrases, set account2 and a transaction comment as shown if monthly service fee atm transaction fee banking thru software account2 expenses:business:banking comment XXX deductible ? check it # if an empty record is seen (assuming five fields), ignore the rest of the CSV file if ,,,, end Matchers There are two kinds of matcher: 1. A whole record matcher is simplest: it is just a word, single-line text fragment, or other regular expression, which hledger will try to match case-insensitively anywhere within the CSV record. Eg: whole foods. 2. A field matcher has a percent-prefixed CSV field number or name be- fore the pattern. Eg: %3 whole foods or %description whole foods. hledger will try to match the pattern just within the named CSV field. When using these, there's two things to be aware of: 1. Whole record matchers don't see the exact original record; they see a reconstruction of it, in which values are comma-separated, and quotes enclosing values and whitespace outside those quotes are re- moved. Eg when reading an SSV record like: 2023-01-01 ; "Acme, Inc. " ; 1,000 the whole record matcher sees instead: 2023-01-01,Acme, Inc. ,1,000 2. Field matchers expect either a CSV field number, or a CSV field name declared with fields. (Don't use a hledger field name here, unless it is also a CSV field name.) A non-CSV field name will cause the matcher to match against "" (the empty string), and does not raise an error, allowing easier reuse of common rules with different CSV files. You can also prefix a matcher with ! (and optional space) to negate it. Eg ! whole foods, ! %3 whole foods, !%description whole foods will match if "whole foods" is NOT present. Added in 1.32. The pattern is, as usual in hledger, a POSIX extended regular expres- sion that also supports GNU word boundaries (\b, \B, \<, \>) and noth- ing else. If you have trouble with it, see "Regular expressions" in the hledger manual (https://hledger.org/hledger.html#regular-expres- sions). Multiple matchers When an if block has multiple matchers, each on its own line, o By default they are OR'd (any of them can match). o Matcher lines beginning with & (or &&, since 1.42) are AND'ed with the matcher above (all in the AND'ed group must match). o Matcher lines beginning with & ! (since 1.41, or && !, since 1.42) are first negated and then AND'ed with the matcher above. You can also combine multiple matchers one the same line separated by && (AND) or && ! (AND NOT). Eg %description amazon && %date 2025-01-01 will match only when the description field contains "amazon" and the date field contains "2025-01-01". Added in 1.42. Match groups Added in 1.32 Matchers can define match groups: parenthesised portions of the regular expression which are available for reference in field assignments. Groups are enclosed in regular parentheses (( and )) and can be nested. Each group is available in field assignments using the token \N, where N is an index into the match groups for this conditional block (e.g. \1, \2, etc.). Example: Warp credit card payment postings to the beginning of the billing period (Month start), to match how they are presented in state- ments, using posting dates: if %date (....-..)-.. comment2 date:\1-01 Another example: Read the expense account from the CSV field, but throw away a prefix: if %account1 liabilities:family:(expenses:.*) account1 \1 if table "if tables" are an alternative to if blocks; they can express many matchers and field assignments in a more compact tabular format, like this: if,HLEDGERFIELD1,HLEDGERFIELD2,... MATCHERA,VALUE1,VALUE2,... MATCHERB && MATCHERC,VALUE1,VALUE2,... (*since 1.42*) ; Comment line that explains MATCHERD MATCHERD,VALUE1,VALUE2,... The first character after if is taken to be this if table's field sepa- rator. It is unrelated to the separator used in the CSV file. It should be a non-alphanumeric character like , or | that does not appear anywhere else in the table (it should not be used in field names or matchers or values, and it cannot be escaped with a backslash). Each line must contain the same number of separators; empty values are allowed. Whitespace can be used in the matcher lines for readability (but not in the if line, currently). You can use the comment lines in the table body. The table must be terminated by an empty line (or end of file). An if table like the above is interpreted as follows: try all of the lines with matchers; whenever a line with matchers succeeds, assign all of the values on that line to the corresponding hledger fields; If mul- tiple lines match, later lines will override fields assigned by the earlier ones - just like the sequence of if blocks would behave. If table presented above is equivalent to this sequence of if blocks: if MATCHERA HLEDGERFIELD1 VALUE1 HLEDGERFIELD2 VALUE2 ... if MATCHERB && MATCHERC HLEDGERFIELD1 VALUE1 HLEDGERFIELD2 VALUE2 ... ; Comment line which explains MATCHERD if MATCHERD HLEDGERFIELD1 VALUE1 HLEDGERFIELD2 VALUE2 ... Example: if,account2,comment atm transaction fee,expenses:business:banking,deductible? check it %description groceries,expenses:groceries, ;; Comment line that desribes why this particular date is special 2023/01/12.*Plumbing LLC,expenses:house:upkeep,emergency plumbing call-out balance-type Balance assertions generated by assigning to balanceN are of the simple = type by default, which is a single-commodity, subaccount-excluding assertion. You may find the subaccount-including variants more useful, eg if you have created some virtual subaccounts of checking to help with budgeting. You can select a different type of assertion with the balance-type rule: # balance assertions will consider all commodities and all subaccounts balance-type ==* Here are the balance assertion types for quick reference: = single commodity, exclude subaccounts =* single commodity, include subaccounts == multi commodity, exclude subaccounts ==* multi commodity, include subaccounts include include RULESFILE This includes the contents of another CSV rules file at this point. RULESFILE is an absolute file path or a path relative to the current file's directory. This can be useful for sharing common rules between several rules files, eg: # someaccount.csv.rules ## someaccount-specific rules fields date,description,amount account1 assets:someaccount account2 expenses:misc ## common rules include categorisation.rules Working with CSV Some tips: Rapid feedback It's a good idea to get rapid feedback while creating/troubleshooting CSV rules. Here's a good way, using entr from eradman.com/entrproject: $ ls foo.csv* | entr bash -c 'echo ----; hledger -f foo.csv print desc:SOMEDESC' A desc: query (eg) is used to select just one, or a few, transactions of interest. "bash -c" is used to run multiple commands, so we can echo a separator each time the command re-runs, making it easier to read the output. Valid CSV Note that hledger will only accept valid CSV conforming to RFC 4180, and equivalent SSV and TSV formats (like RFC 4180 but with semicolon or tab as separators). This means, eg: o Values may be enclosed in double quotes, or not. Enclosing in single quotes is not allowed. (Eg 'A','B' is rejected.) o When values are enclosed in double quotes, spaces outside the quotes are not allowed. (Eg "A", "B" is rejected.) o When values are not enclosed in quotes, they may not contain double quotes. (Eg A"A, B is rejected.) If your CSV/SSV/TSV is not valid in this sense, you'll need to trans- form it before reading with hledger. Try using sed, or a more permis- sive CSV parser like python's csv lib. File Extension To help hledger choose the CSV file reader and show the right error messages (and choose the right field separator character by default), it's best if CSV/SSV/TSV files are named with a .csv, .ssv or .tsv filename extension. (More about this at Data formats.) When reading files with the "wrong" extension, you can ensure the CSV reader (and the default field separator) by prefixing the file path with csv:, ssv: or tsv:: Eg: $ hledger -f ssv:foo.dat print You can also override the default field separator with a separator rule if needed. Reading CSV from standard input You'll need the file format prefix when reading CSV from stdin also, since hledger assumes journal format by default. Eg: $ cat foo.dat | hledger -f ssv:- print Reading multiple CSV files If you use multiple -f options to read multiple CSV files at once, hledger will look for a correspondingly-named rules file for each CSV file. But if you specify a rules file with --rules, that rules file will be used for all the CSV files. Reading files specified by rule Instead of specifying a CSV file in the command line, you can specify a rules file, as in hledger -f foo.csv.rules CMD. By default this will read data from foo.csv in the same directory, but you can add a source rule to specify a different data file, perhaps located in your web browser's download directory. This feature was added in hledger 1.30, so you won't see it in most CSV rules examples. But it helps remove some of the busywork of managing CSV downloads. Most of your financial institutions's default CSV file- names are different and can be recognised by a glob pattern. So you can put a rule like source Checking1*.csv in foo-checking.csv.rules, and then periodically follow a workflow like: 1. Download CSV from Foo's website, using your browser's defaults 2. Run hledger import foo-checking.csv.rules to import any new transac- tions After import, you can: discard the CSV, or leave it where it is for a while, or move it into your archives, as you prefer. If you do noth- ing, next time your browser will save something like Checking1-2.csv, and hledger will use that because of the * wild card and because it is the most recent. Valid transactions After reading a CSV file, hledger post-processes and validates the gen- erated journal entries as it would for a journal file - balancing them, applying balance assignments, and canonicalising amount styles. Any errors at this stage will be reported in the usual way, displaying the problem entry. There is one exception: balance assertions, if you have generated them, will not be checked, since normally these will work only when the CSV data is part of the main journal. If you do need to check balance as- sertions generated from CSV right away, pipe into another hledger: $ hledger -f file.csv print | hledger -f- print Deduplicating, importing When you download a CSV file periodically, eg to get your latest bank transactions, the new file may overlap with the old one, containing some of the same records. The import command will (a) detect the new transactions, and (b) append just those transactions to your main journal. It is idempotent, so you don't have to remember how many times you ran it or with which version of the CSV. (It keeps state in a hidden .latest.FILE.csv file.) This is the easiest way to import CSV data. Eg: # download the latest CSV files, then run this command. # Note, no -f flags needed here. $ hledger import *.csv [--dry] This method works for most CSV files. (Where records have a stable chronological order, and new records appear only at the new end.) A number of other tools and workflows, hledger-specific and otherwise, exist for converting, deduplicating, classifying and managing CSV data. See: o https://hledger.org/cookbook.html#setups-and-workflows o https://plaintextaccounting.org -> data import/conversion Setting amounts Continuing from amount field above, here are more tips for amount-set- ting: 1. If the amount is in a single CSV field: a. If its sign indicates direction of flow: Assign it to amountN, to set the Nth posting's amount. N is usu- ally 1 or 2 but can go up to 99. b. If another field indicates direction of flow: Use one or more conditional rules to set the appropriate amount sign. Eg: # assume a withdrawal unless Type contains "deposit": amount1 -%Amount if %Type deposit amount1 %Amount 2. If the amount is in two CSV fields (such as Debit and Credit, or In and Out): a. If both fields are unsigned: Assign one field to amountN-in and the other to amountN-out. hledger will automatically negate the "out" field, and will use whichever field value is non-zero as posting N's amount. b. If either field is signed: You will probably need to override hledger's sign for one or the other field, as in the following example: # Negate the -out value, but only if it is not empty: fields date, description, amount1-in, amount1-out if %amount1-out [1-9] amount1-out -%amount1-out c. If both fields can contain a non-zero value (or both can be empty): The -in/-out rules normally choose the value which is non-zero/non-empty. Some value pairs can be ambiguous, such as 1 and none. For such cases, use conditional rules to help select the amount. Eg, to handle the above you could select the value con- taining non-zero digits: fields date, description, in, out if %in [1-9] amount1 %in if %out [1-9] amount1 %out 3. If you want posting 2's amount converted to cost: Use the unnumbered amount (or amount-in and amount-out) syntax. 4. If the CSV has only balance amounts, not transaction amounts: Assign to balanceN, to set a balance assignment on the Nth posting, causing the posting's amount to be calculated automatically. balance with no number is equivalent to balance1. In this situation hledger is more likely to guess the wrong default account name, so you may need to set that explicitly. Amount signs There is some special handling making it easier to parse and to reverse amount signs. (This only works for whole amounts, not for cost amounts such as COST in amount1 AMT @ COST): o If an amount value begins with a plus sign: that will be removed: +AMT becomes AMT o If an amount value is parenthesised: it will be de-parenthesised and sign-flipped: (AMT) becomes -AMT o If an amount value has two minus signs (or two sets of parentheses, or a minus sign and parentheses): they cancel out and will be removed: --AMT or -(AMT) becomes AMT o If an amount value contains just a sign (or just a set of parenthe- ses): that is removed, making it an empty value. "+" or "-" or "()" becomes "". It's not possible (without preprocessing the CSV) to set an amount to its absolute value, ie discard its sign. Setting currency/commodity If the currency/commodity symbol is included in the CSV's amount field(s): 2023-01-01,foo,$123.00 you don't have to do anything special for the commodity symbol, it will be assigned as part of the amount. Eg: fields date,description,amount 2023-01-01 foo expenses:unknown $123.00 income:unknown $-123.00 If the currency is provided as a separate CSV field: 2023-01-01,foo,USD,123.00 You can assign that to the currency pseudo-field, which has the special effect of prepending itself to every amount in the transaction (on the left, with no separating space): fields date,description,currency,amount 2023-01-01 foo expenses:unknown USD123.00 income:unknown USD-123.00 Or, you can use a field assignment to construct the amount yourself, with more control. Eg to put the symbol on the right, and separated by a space: fields date,description,cur,amt amount %amt %cur 2023-01-01 foo expenses:unknown 123.00 USD income:unknown -123.00 USD Note we used a temporary field name (cur) that is not currency - that would trigger the prepending effect, which we don't want here. Amount decimal places When you are reading CSV data, eg with a command like hledger -f foo.csv print, hledger will infer each commodity's decimal precision (and other commodity display styles) from the amounts - much as when reading a journal file without commodity directives (see the link). Note, the commodity styles are not inferred from the numbers in the original CSV data; rather, they are inferred from the amounts generated by the CSV rules. When you are importing CSV data with the import command, eg hledger im- port foo.csv, there's another step: import tries to make the new en- tries conform to the journal's existing styles. So for each commodity - let's say it's EUR - import will choose: 1. the style declared for EUR by a commodity directive in the journal 2. otherwise, the style inferred from EUR amounts in the journal 3. otherwise, the style inferred from EUR amounts generated by the CSV rules. TLDR: if import is not generating the precisions or styles you want, add a commodity directive to specify them. Referencing other fields In field assignments, you can interpolate only CSV fields, not hledger fields. In the example below, there's both a CSV field and a hledger field named amount1, but %amount1 always means the CSV field, not the hledger field: # Name the third CSV field "amount1" fields date,description,amount1 # Set hledger's amount1 to the CSV amount1 field followed by USD amount1 %amount1 USD # Set comment to the CSV amount1 (not the amount1 assigned above) comment %amount1 Here, since there's no CSV amount1 field, %amount1 will produce a lit- eral "amount1": fields date,description,csvamount amount1 %csvamount USD # Can't interpolate amount1 here comment %amount1 When there are multiple field assignments to the same hledger field, only the last one takes effect. Here, comment's value will be be B, or C if "something" is matched, but never A: comment A comment B if something comment C How CSV rules are evaluated Here's how to think of CSV rules being evaluated. If you get a confus- ing error while reading a CSV file, it may help to try to understand which of these steps is failing: 1. Any included rules files are inlined, from top to bottom, depth first (scanning each included file for further includes, recur- sively, before proceeding). 2. Top level rules (date-format, fields, newest-first, skip etc) are read, top to bottom. "Top level rules" means non-conditional rules. If a rule occurs more than once, the last one wins; except for skip/end rules, where the first one wins. 3. The CSV file is read as text. Any non-ascii characters will be de- coded using the text encoding specified by the encoding rule, other- wise the system locale's text encoding. 4. Any top-level skip or end rule is applied. skip [N] immediately skips the current or next N CSV records; end immediately skips all remaining CSV records (not normally used at top level). 5. Now any remaining CSV records are processed. For each CSV record, in file order: o Is there a conditional skip/end rule that applies for this record ? Search the if blocks, from top to bottom, for a succeeding one containing a skip or end rule. If found, skip the specified num- ber of CSV records, then continue at 5. Otherwise... o Do some basic validation on this CSV record (eg, check that it has at least two fields). o For each hledger field (date, description, account1, etc.): 1. Get the field's assigned value, first searching top level as- signments, made directly or by the fields rule, then assign- ments made inside succeeding if blocks. If there are more than one, the last one wins. 2. Compute the field's actual value (as text), by interpolating any %CSVFIELD references within the assigned value; or by choosing a default value if there was no assignment. o Generate a hledger transaction from the hledger field values, parsing them if needed (eg from text to an amount). This is all done by the CSV reader, one of several readers hledger can use to read transactions from an input file. When all input files have been read successfully, their transactions are passed to whichever hledger command the user specified. Well factored rules Some things than can help reduce duplication and complexity in rules files: o Extracting common rules usable with multiple CSV files into a com- mon.rules, and adding include common.rules to each CSV's rules file. o Splitting if blocks into smaller if blocks, extracting the frequently used parts. CSV rules examples Bank of Ireland Here's a CSV with two amount fields (Debit and Credit), and a balance field, which we can use to add balance assertions, which is not neces- sary but provides extra error checking: Date,Details,Debit,Credit,Balance 07/12/2012,LODGMENT 529898,,10.0,131.21 07/12/2012,PAYMENT,5,,126 # bankofireland-checking.csv.rules # skip the header line skip # name the csv fields, and assign some of them as journal entry fields fields date, description, amount-out, amount-in, balance # We generate balance assertions by assigning to "balance" # above, but you may sometimes need to remove these because: # # - the CSV balance differs from the true balance, # by up to 0.0000000000005 in my experience # # - it is sometimes calculated based on non-chronological ordering, # eg when multiple transactions clear on the same day # date is in UK/Ireland format date-format %d/%m/%Y # set the currency currency EUR # set the base account for all txns account1 assets:bank:boi:checking $ hledger -f bankofireland-checking.csv print 2012-12-07 LODGMENT 529898 assets:bank:boi:checking EUR10.0 = EUR131.2 income:unknown EUR-10.0 2012-12-07 PAYMENT assets:bank:boi:checking EUR-5.0 = EUR126.0 expenses:unknown EUR5.0 The balance assertions don't raise an error above, because we're read- ing directly from CSV, but they will be checked if these entries are imported into a journal file. Coinbase A simple example with some CSV from Coinbase. The spot price is recorded using cost notation. The legacy amount field name conve- niently sets amount 2 (posting 2's amount) to the total cost. # Timestamp,Transaction Type,Asset,Quantity Transacted,Spot Price Currency,Spot Price at Transaction,Subtotal,Total (inclusive of fees and/or spread),Fees and/or Spread,Notes # 2021-12-30T06:57:59Z,Receive,USDC,100,GBP,0.740000,"","","","Received 100.00 USDC from an external account" # coinbase.csv.rules skip 1 fields Timestamp,Transaction_Type,Asset,Quantity_Transacted,Spot_Price_Currency,Spot_Price_at_Transaction,Subtotal,Total,Fees_Spread,Notes date %Timestamp date-format %Y-%m-%dT%T%Z description %Notes account1 assets:coinbase:cc amount %Quantity_Transacted %Asset @ %Spot_Price_at_Transaction %Spot_Price_Currency $ hledger print -f coinbase.csv 2021-12-30 Received 100.00 USDC from an external account assets:coinbase:cc 100 USDC @ 0.740000 GBP income:unknown -74.000000 GBP Amazon Here we convert amazon.com order history, and use an if block to gener- ate a third posting if there's a fee. (In practice you'd probably get this data from your bank instead, but it's an example.) "Date","Type","To/From","Name","Status","Amount","Fees","Transaction ID" "Jul 29, 2012","Payment","To","Foo.","Completed","$20.00","$0.00","16000000000000DGLNJPI1P9B8DKPVHL" "Jul 30, 2012","Payment","To","Adapteva, Inc.","Completed","$25.00","$1.00","17LA58JSKRD4HDGLNJPI1P9B8DKPVHL" # amazon-orders.csv.rules # skip one header line skip 1 # name the csv fields, and assign the transaction's date, amount and code. # Avoided the "status" and "amount" hledger field names to prevent confusion. fields date, _, toorfrom, name, amzstatus, amzamount, fees, code # how to parse the date date-format %b %-d, %Y # combine two fields to make the description description %toorfrom %name # save the status as a tag comment status:%amzstatus # set the base account for all transactions account1 assets:amazon # leave amount1 blank so it can balance the other(s). # I'm assuming amzamount excludes the fees, don't remember # set a generic account2 account2 expenses:misc amount2 %amzamount # and maybe refine it further: #include categorisation.rules # add a third posting for fees, but only if they are non-zero. if %fees [1-9] account3 expenses:fees amount3 %fees $ hledger -f amazon-orders.csv print 2012-07-29 (16000000000000DGLNJPI1P9B8DKPVHL) To Foo. ; status:Completed assets:amazon expenses:misc $20.00 2012-07-30 (17LA58JSKRD4HDGLNJPI1P9B8DKPVHL) To Adapteva, Inc. ; status:Completed assets:amazon expenses:misc $25.00 expenses:fees $1.00 Paypal Here's a real-world rules file for (customised) Paypal CSV, with some Paypal-specific rules, and a second rules file included: "Date","Time","TimeZone","Name","Type","Status","Currency","Gross","Fee","Net","From Email Address","To Email Address","Transaction ID","Item Title","Item ID","Reference Txn ID","Receipt ID","Balance","Note" "10/01/2019","03:46:20","PDT","Calm Radio","Subscription Payment","Completed","USD","-6.99","0.00","-6.99","simon@joyful.com","memberships@calmradio.com","60P57143A8206782E","MONTHLY - $1 for the first 2 Months: Me - Order 99309. Item total: $1.00 USD first 2 months, then $6.99 / Month","","I-R8YLY094FJYR","","-6.99","" "10/01/2019","03:46:20","PDT","","Bank Deposit to PP Account ","Pending","USD","6.99","0.00","6.99","","simon@joyful.com","0TU1544T080463733","","","60P57143A8206782E","","0.00","" "10/01/2019","08:57:01","PDT","Patreon","PreApproved Payment Bill User Payment","Completed","USD","-7.00","0.00","-7.00","simon@joyful.com","support@patreon.com","2722394R5F586712G","Patreon* Membership","","B-0PG93074E7M86381M","","-7.00","" "10/01/2019","08:57:01","PDT","","Bank Deposit to PP Account ","Pending","USD","7.00","0.00","7.00","","simon@joyful.com","71854087RG994194F","Patreon* Membership","","2722394R5F586712G","","0.00","" "10/19/2019","03:02:12","PDT","Wikimedia Foundation, Inc.","Subscription Payment","Completed","USD","-2.00","0.00","-2.00","simon@joyful.com","tle@wikimedia.org","K9U43044RY432050M","Monthly donation to the Wikimedia Foundation","","I-R5C3YUS3285L","","-2.00","" "10/19/2019","03:02:12","PDT","","Bank Deposit to PP Account ","Pending","USD","2.00","0.00","2.00","","simon@joyful.com","3XJ107139A851061F","","","K9U43044RY432050M","","0.00","" "10/22/2019","05:07:06","PDT","Noble Benefactor","Subscription Payment","Completed","USD","10.00","-0.59","9.41","noble@bene.fac.tor","simon@joyful.com","6L8L1662YP1334033","Joyful Systems","","I-KC9VBGY2GWDB","","9.41","" # paypal-custom.csv.rules # Tips: # Export from Activity -> Statements -> Custom -> Activity download # Suggested transaction type: "Balance affecting" # Paypal's default fields in 2018 were: # "Date","Time","TimeZone","Name","Type","Status","Currency","Gross","Fee","Net","From Email Address","To Email Address","Transaction ID","Shipping Address","Address Status","Item Title","Item ID","Shipping and Handling Amount","Insurance Amount","Sales Tax","Option 1 Name","Option 1 Value","Option 2 Name","Option 2 Value","Reference Txn ID","Invoice Number","Custom Number","Quantity","Receipt ID","Balance","Address Line 1","Address Line 2/District/Neighborhood","Town/City","State/Province/Region/County/Territory/Prefecture/Republic","Zip/Postal Code","Country","Contact Phone Number","Subject","Note","Country Code","Balance Impact" # This rules file assumes the following more detailed fields, configured in "Customize report fields": # "Date","Time","TimeZone","Name","Type","Status","Currency","Gross","Fee","Net","From Email Address","To Email Address","Transaction ID","Item Title","Item ID","Reference Txn ID","Receipt ID","Balance","Note" fields date, time, timezone, description_, type, status_, currency, grossamount, feeamount, netamount, fromemail, toemail, code, itemtitle, itemid, referencetxnid, receiptid, balance, note skip 1 date-format %-m/%-d/%Y # ignore some paypal events if In Progress Temporary Hold Update to skip # add more fields to the description description %description_ %itemtitle # save some other fields as tags comment itemid:%itemid, fromemail:%fromemail, toemail:%toemail, time:%time, type:%type, status:%status_ # convert to short currency symbols if %currency USD currency $ if %currency EUR currency E if %currency GBP currency P # generate postings # the first posting will be the money leaving/entering my paypal account # (negative means leaving my account, in all amount fields) account1 assets:online:paypal amount1 %netamount # the second posting will be money sent to/received from other party # (account2 is set below) amount2 -%grossamount # if there's a fee, add a third posting for the money taken by paypal. if %feeamount [1-9] account3 expenses:banking:paypal amount3 -%feeamount comment3 business: # choose an account for the second posting # override the default account names: # if the amount is positive, it's income (a debit) if %grossamount ^[^-] account2 income:unknown # if negative, it's an expense (a credit) if %grossamount ^- account2 expenses:unknown # apply common rules for setting account2 & other tweaks include common.rules # apply some overrides specific to this csv # Transfers from/to bank. These are usually marked Pending, # which can be disregarded in this case. if Bank Account Bank Deposit to PP Account description %type for %referencetxnid %itemtitle account2 assets:bank:wf:pchecking account1 assets:online:paypal # Currency conversions if Currency Conversion account2 equity:currency conversion # common.rules if darcs noble benefactor account2 revenues:foss donations:darcshub comment2 business: if Calm Radio account2 expenses:online:apps if electronic frontier foundation Patreon wikimedia Advent of Code account2 expenses:dues if Google account2 expenses:online:apps description google | music $ hledger -f paypal-custom.csv print 2019-10-01 (60P57143A8206782E) Calm Radio MONTHLY - $1 for the first 2 Months: Me - Order 99309. Item total: $1.00 USD first 2 months, then $6.99 / Month ; itemid:, fromemail:simon@joyful.com, toemail:memberships@calmradio.com, time:03:46:20, type:Subscription Payment, status:Completed assets:online:paypal $-6.99 = $-6.99 expenses:online:apps $6.99 2019-10-01 (0TU1544T080463733) Bank Deposit to PP Account for 60P57143A8206782E ; itemid:, fromemail:, toemail:simon@joyful.com, time:03:46:20, type:Bank Deposit to PP Account, status:Pending assets:online:paypal $6.99 = $0.00 assets:bank:wf:pchecking $-6.99 2019-10-01 (2722394R5F586712G) Patreon Patreon* Membership ; itemid:, fromemail:simon@joyful.com, toemail:support@patreon.com, time:08:57:01, type:PreApproved Payment Bill User Payment, status:Completed assets:online:paypal $-7.00 = $-7.00 expenses:dues $7.00 2019-10-01 (71854087RG994194F) Bank Deposit to PP Account for 2722394R5F586712G Patreon* Membership ; itemid:, fromemail:, toemail:simon@joyful.com, time:08:57:01, type:Bank Deposit to PP Account, status:Pending assets:online:paypal $7.00 = $0.00 assets:bank:wf:pchecking $-7.00 2019-10-19 (K9U43044RY432050M) Wikimedia Foundation, Inc. Monthly donation to the Wikimedia Foundation ; itemid:, fromemail:simon@joyful.com, toemail:tle@wikimedia.org, time:03:02:12, type:Subscription Payment, status:Completed assets:online:paypal $-2.00 = $-2.00 expenses:dues $2.00 expenses:banking:paypal ; business: 2019-10-19 (3XJ107139A851061F) Bank Deposit to PP Account for K9U43044RY432050M ; itemid:, fromemail:, toemail:simon@joyful.com, time:03:02:12, type:Bank Deposit to PP Account, status:Pending assets:online:paypal $2.00 = $0.00 assets:bank:wf:pchecking $-2.00 2019-10-22 (6L8L1662YP1334033) Noble Benefactor Joyful Systems ; itemid:, fromemail:noble@bene.fac.tor, toemail:simon@joyful.com, time:05:07:06, type:Subscription Payment, status:Completed assets:online:paypal $9.41 = $9.41 revenues:foss donations:darcshub $-10.00 ; business: expenses:banking:paypal $0.59 ; business: Timeclock hledger can read time logs in the timeclock time logging format of timeclock.el. As with Ledger, hledger's timeclock format is a sub- set/variant of timeclock.el's. hledger's timeclock format was updated in hledger 1.43 and 1.50. If your old time logs are rejected, you should adapt them to modern hledger; for now, you can restore the pre-1.43 behaviour with the --old-timeclock flag. Here the timeclock format in hledger 1.50+: # Comment lines like these, and blank lines, are ignored: # comment line ; comment line * comment line # Lines beginning with b, h, or capital O are also ignored, for compatibility: b SIMPLEDATE HH:MM[:SS][+-ZZZZ][ TEXT] h SIMPLEDATE HH:MM[:SS][+-ZZZZ][ TEXT] O SIMPLEDATE HH:MM[:SS][+-ZZZZ][ TEXT] # Lines beginning with i or o are are clock-in / clock-out entries: i SIMPLEDATE HH:MM[:SS][+-ZZZZ] ACCOUNT[ DESCRIPTION][;COMMENT]] o SIMPLEDATE HH:MM[:SS][+-ZZZZ][ ACCOUNT][;COMMENT] The date is a hledger simple date (YYYY-MM-DD or similar). The time parts must use two digits. The seconds are optional. A + or - four-digit time zone is accepted for compatibility, but currently ig- nored; times are always interpreted as a local time. In clock-in entries (i), the account name is required. A transaction description, separated from the account name by 2+ spaces, is optional. A transaction comment, beginning with ;, is also optional. (Indented following comment lines are also allowed, as in journal format.) In clock-out entries (o) have no description, but can have a comment if you wish. A clock-in and clock-out pair form a "transaction" posting some number of hours to an account - also known as a session. Eg: i 2015/03/30 09:00:00 session1 o 2015/03/30 10:00:00 $ hledger -f a.timeclock print 2015-03-30 * 09:00-10:00 (session1) 1.00h Clock-ins and clock-outs are matched by their account/session name. If a clock-out does not specify a name, the most recent unclosed clock-in is closed. You can have multiple sessions active simultaneously. En- tries are processed in the order they are parsed. Sessions spanning more than one day are automatically split at day boundaries. Eg, the following time log: i 2015/03/30 09:00:00 some account optional description after 2 spaces ; optional comment, tags: o 2015/03/30 09:20:00 i 2015/03/31 22:21:45 another:account o 2015/04/01 02:00:34 i 2015/04/02 12:00:00 another:account ; this demonstrates multple sessions being clocked in i 2015/04/02 13:00:00 some account o 2015/04/02 14:00:00 o 2015/04/02 15:00:00 another:account generates these transactions: $ hledger -f t.timeclock print 2015-03-30 * optional description after 2 spaces ; optional comment, tags: (some account) 0.33h 2015-03-31 * 22:21-23:59 (another:account) 1.64h 2015-04-01 * 00:00-02:00 (another:account) 2.01h 2015-04-02 * 12:00-15:00 ; this demonstrates multiple sessions being clocked in (another:account) 3.00h 2015-04-02 * 13:00-14:00 (some account) 1.00h Here is a sample.timeclock to download and some queries to try: $ hledger -f sample.timeclock balance # current time balances $ hledger -f sample.timeclock register -p 2009/3 # sessions in march 2009 $ hledger -f sample.timeclock register -p weekly --depth 1 --empty # time summary by week To generate time logs, ie to clock in and clock out, you could: o use these shell aliases at the command line: alias ti='echo i `date "+%Y-%m-%d %H:%M:%S"` $* >>$TIMELOG' alias to='echo o `date "+%Y-%m-%d %H:%M:%S"` >>$TIMELOG' o or Emacs's built-in timeclock.el, or the extended timeclock-x.el, and perhaps the extras in ledgerutils.el o or use the old ti and to scripts in the ledger 2.x repository. These rely on a "timeclock" executable which I think is just the ledger 2 executable renamed. Timedot timedot format is hledger's human-friendly time logging format. Com- pared to timeclock format, it is more convenient for quick, approxi- mate, and retroactive time logging, and more human-readable (you can see at a glance where time was spent). A quick example: 2023-05-01 hom:errands .... .... ; two hours; the space is ignored fos:hledger:timedot .. ; half an hour per:admin:finance ; no time spent yet hledger reads this as a transaction on this day with three (unbalanced) postings, where each dot represents "0.25". No commodity symbol is as- sumed, but we typically interpret it as hours. $ hledger -f a.timedot print # .timedot file extension (or timedot: prefix) is required 2023-05-01 * (hom:errands) 2.00 ; two hours (fos:hledger:timedot) 0.50 ; half an hour (per:admin:finance) 0 A timedot file contains a series of transactions (usually one per day). Each begins with a simple date (Y-M-D, Y/M/D, or Y.M.D), optionally be followed on the same line by a transaction description, and/or a trans- action comment following a semicolon. After the date line are zero or more time postings, consisting of: o An account name - any hledger-style account name, optionally in- dented. o Two or more spaces - required if there is an amount (as in journal format). o A timedot amount, which can be o empty (representing zero) o a number, optionally followed by a unit s, m, h, d, w, mo, or y, representing a precise number of seconds, minutes, hours, days weeks, months or years (hours is assumed by default), which will be converted to hours according to 60s = 1m, 60m = 1h, 24h = 1d, 7d = 1w, 30d = 1mo, 365d = 1y. o one or more dots (period characters), each representing 0.25. These are the dots in "timedot". Spaces are ignored and can be used for grouping/alignment. o Added in 1.32 one or more letters. These are like dots but they also generate a tag t: (short for "type") with the letter as its value, and a separate posting for each of the values. This pro- vides a second dimension of categorisation, viewable in reports with --pivot t. o An optional comment following a semicolon (a hledger-style posting comment). There is some flexibility to help with keeping time log data and notes in the same file: o Blank lines and lines beginning with # or ; are ignored. o After the first date line, lines which do not contain a double space are parsed as postings with zero amount. (hledger's register reports will show these if you add -E). o Before the first date line, lines beginning with * (eg org headings) are ignored. And from the first date line onward, Emacs org mode heading prefixes at the start of lines (one or more *'s followed by a space) will be ignored. This means the time log can also be a org outline. Timedot files don't support directives like journal files. So a common pattern is to have a main journal file (eg time.journal) that contains any needed directives, and then includes the timedot file (include time.timedot). Timedot examples Numbers: 2016/2/3 inc:client1 4 fos:hledger 3h biz:research 60m Dots: # on this day, 6h was spent on client work, 1.5h on haskell FOSS work, etc. 2016/2/1 inc:client1 .... .... .... .... .... .... fos:haskell .... .. biz:research . 2016/2/2 inc:client1 .... .... biz:research . $ hledger -f a.timedot print date:2016/2/2 2016-02-02 * (inc:client1) 2.00 2016-02-02 * (biz:research) 0.25 $ hledger -f a.timedot bal --daily --tree Balance changes in 2016-02-01-2016-02-03: || 2016-02-01d 2016-02-02d 2016-02-03d ============++======================================== biz || 0.25 0.25 1.00 research || 0.25 0.25 1.00 fos || 1.50 0 3.00 haskell || 1.50 0 0 hledger || 0 0 3.00 inc || 6.00 2.00 4.00 client1 || 6.00 2.00 4.00 ------------++---------------------------------------- || 7.75 2.25 8.00 Letters: # Activity types: # c cleanup/catchup/repair # e enhancement # s support # l learning/research 2023-11-01 work:adm ccecces $ hledger -f a.timedot print 2023-11-01 (work:adm) 1 ; t:c (work:adm) 0.5 ; t:e (work:adm) 0.25 ; t:s $ hledger -f a.timedot bal 1.75 work:adm -------------------- 1.75 $ hledger -f a.timedot bal --pivot t 1.00 c 0.50 e 0.25 s -------------------- 1.75 Org: * 2023 Work Diary ** Q1 *** 2023-02-29 **** DONE 0700 yoga **** UNPLANNED **** BEGUN hom:chores cleaning ... water plants outdoor - one full watering can indoor - light watering **** TODO adm:planning: trip *** LATER Using . as account name separator: 2016/2/4 fos.hledger.timedot 4h fos.ledger .. $ hledger -f a.timedot --alias '/\./=:' bal -t 4.50 fos 4.00 hledger:timedot 0.50 ledger -------------------- 4.50 PART 3: REPORTING CONCEPTS Time periods Report start & end date Most hledger reports will by default show the full time period repre- sented by the journal. The report start date will be the earliest transaction or posting date, and the report end date will be the latest transaction, posting, or market price date. Often you will want to see a shorter period, such as the current month. You can specify a start and/or end date with the -b/--begin, -e/--end, or -p/--period options, or a date: query argument, described below. All of these accept the smart date syntax, also described below. End dates are exclusive; specify the day after the last day you want to see in the report. When dates are specified by multiple options, the last (right-most) op- tion wins. And when date: queries and date options are combined, the report period will be their intersection. Examples: -b 2016/3/17 beginning on St. Patrick's day 2016 -e 12/1 ending at the start of December 1st in the current year -p 'this month' during the current month -p thismonth same as above, spaces are optional -b 2023 beginning on the first day of 2023 date:2023.. or date:2023- same as above -b 2024 -e 2025 -p '2000 to 2030' date:2020-01 date:2020 : during January 2020 (the smallest common period, with the -p overriding -b and -e) Smart dates In hledger's user interfaces (though not in the journal file), you can optionally use "smart date" syntax. Smart dates can be written with english words, can be relative, and can have parts omitted. Missing parts are inferred as 1, when needed. Smart dates can be interpreted as dates or periods depending on the context. Examples: 2004-01-01, 2004/10/1, 2004.9.1, 20240504, 2024Q1 : Exact dates. The year must have at least four digits, the month must be 1-12, the day must be 1-31, the separator can be - or / or . or nothing. The q can be upper or lower case and the quarter number must be 1-4. 2004-10 start of month 2004q3 start of third quarter of 2004 q3 start of third quarter of current year 2004 start of year 10/1 or oct or october October 1st in current year 21 21st day in current month yesterday, today, tomorrow -1, 0, 1 days from today last/this/next day/week/month/quarter/year -1, 0, 1 periods from the current period in n days/weeks/months/quarters/years n periods from the current period n days/weeks/months/quarters/years ahead n periods from the current period n days/weeks/months/quarters/years ago -n periods from the current period 20181201 8 digit YYYYMMDD with valid year month and day 201812 6 digit YYYYMM with valid year and month Dates with no separators are allowed but might give surprising results if mistyped: o 20181301 (YYYYMMDD with an invalid month) is parsed as an eight-digit year o 20181232 (YYYYMMDD with an invalid day) gives a parse error o 201801012 (a valid YYYYMMDD followed by additional digits) gives a parse error The meaning of relative dates depends on today's date. If you need to test or reproduce old reports, you can use the --today option to over- ride that. (Except for periodic transaction rules, which are not af- fected by --today.) Report intervals A report interval can be specified so that reports like register, bal- ance or activity become multi-period, showing each subperiod as a sepa- rate row or column. The following standard intervals can be enabled with command-line flags: o -D/--daily o -W/--weekly o -M/--monthly o -Q/--quarterly o -Y/--yearly More complex intervals can be specified using -p/--period, described below. Date adjustments Start date adjustment If you let hledger infer a report's start date, it will adjust the date to the previous natural boundary of the report interval, for convenient periodic reports. (If you don't want that, specify a start date.) For example, if the journal's first transaction is on january 10th, o hledger register (no report interval) will start the report on janu- ary 10th. o hledger register --monthly will start the report on the previous month boundary, january 1st. o hledger register --monthly --begin 1/5 will start the report on janu- ary 5th [1]. Also if you are generating transactions or budget goals with periodic transaction rules, their start date may be adjusted in a similar way (in certain situations). End date adjustment A report's end date is always adjusted to include a whole number of in- tervals, so that the last subperiod has the same length as the others. For example, if the journal's last transaction is on february 20th, o hledger register will end the report on february 20th. o hledger register --monthly will end the report at the end of febru- ary. o hledger register --monthly --end 2/14 also will end the report at the end of february (overriding the requested end date). o hledger register --monthly --begin 1/5 --end 2/14 will end the report on march 4th [1]. [1] Since hledger 1.29. Period headings With non-standard subperiods, hledger will show "STARTDATE..ENDDATE" headings. With standard subperiods (ie, starting on a natural interval boundary), you'll see more compact headings, which are usually prefer- able. (Though month names will be in english, currently.) So if you are specifying a start date and you want compact headings: choose a start of year for yearly reports, a start of quarter for quar- terly reports, a start of month for monthly reports, etc. (Remember, you can write eg -b 2024 or 1/1 as a shortcut for a start of year, or 2024-04 or 202404 or Apr for a start of month or quarter.) For weekly reports, choose a date that's a Monday. (You can try dif- ferent dates until you see the short headings, or write eg -b '3 weeks ago'.) Period expressions The -p/--period option specifies a period expression, which is a com- pact way of expressing a start date, end date, and/or report interval. Here's a period expression with a start and end date (specifying the first quarter of 2009): -p "from 2009/1/1 to 2009/4/1" Several keywords like "from" and "to" are supported for readability; these are optional. "to" can also be written as ".." or "-". The spaces are also optional, as long as you don't run two dates together. So the following are equivalent to the above: -p "2009/1/1 2009/4/1" -p2009/1/1to2009/4/1 -p2009/1/1..2009/4/1 Dates are smart dates, so if the current year is 2009, these are also equivalent to the above: -p "1/1 4/1" -p "jan-apr" -p "this year to 4/1" If you specify only one date, the missing start or end date will be the earliest or latest transaction date in the journal: -p "from 2009/1/1" everything after january 1, 2009 -p "since 2009/1" the same, since is a syn- onym -p "from 2009" the same -p "to 2009" everything before january 1, 2009 You can also specify a period by writing a single partial or full date: -p "2009" the year 2009; equivalent to "2009/1/1 to 2010/1/1" -p "2009/1" the month of january 2009; equivalent to "2009/1/1 to 2009/2/1" -p "2009/1/1" the first day of 2009; equivalent to "2009/1/1 to 2009/1/2" or by using the "Q" quarter-year syntax (case insensitive): -p "2009Q1" first quarter of 2009, equivalent to "2009/1/1 to 2009/4/1" -p "q4" fourth quarter of the current year Period expressions with a report interval A period expression can also begin with a report interval, separated from the start/end dates (if any) by a space or the word in: -p "weekly from 2009/1/1 to 2009/4/1" -p "monthly in 2008" -p "quarterly" More complex report intervals Some more complex intervals can be specified within period expressions, such as: o biweekly (every two weeks) o fortnightly o bimonthly (every two months) o every day|week|month|quarter|year o every N days|weeks|months|quarters|years Weekly on a custom day: o every Nth day of week (th, nd, rd, or st are all accepted after the number) o every WEEKDAYNAME (full or three-letter english weekday name, case insensitive) Monthly on a custom day: o every Nth day [of month] (31st day will be adjusted to each month's last day) o every Nth WEEKDAYNAME [of month] Yearly on a custom month and day: o every MM/DD [of year] (month number and day of month number) o every MONTHNAME DDth [of year] (full or three-letter english month name, case insensitive, and day of month number) o every DDth MONTHNAME [of year] (equivalent to the above) Examples: -p "bimonthly from 2008" -p "every 2 weeks" -p "every 5 months from 2009/03" -p "every 2nd day of week" periods will go from Tue to Tue -p "every Tue" same -p "every 15th day" period boundaries will be on 15th of each month -p "every 2nd Monday" period boundaries will be on second Monday of each month -p "every 11/05" yearly periods with boundaries on 5th of November -p "every 5th November" same -p "every Nov 5th" same Show historical balances at end of the 15th day of each month (N is an end date, exclusive as always): $ hledger balance -H -p "every 16th day" Group postings from the start of wednesday to end of the following tuesday (N is both (inclusive) start date and (exclusive) end date): $ hledger register checking -p "every 3rd day of week" Multiple weekday intervals This special form is also supported: o every WEEKDAYNAME,WEEKDAYNAME,... (full or three-letter english week- day names, case insensitive) Also, weekday and weekendday are shorthand for mon,tue,wed,thu,fri and sat,sun. This is mainly intended for use with --forecast, to generate periodic transactions on arbitrary days of the week. It may be less useful with -p, since it divides each week into subperiods of unequal length, which is unusual. (Related: #1632) Examples: -p "every dates will be Mon, Wed, Fri; periods will be mon,wed,fri" Mon-Tue, Wed-Thu, Fri-Sun -p "every weekday" dates will be Mon, Tue, Wed, Thu, Fri; periods will be Mon, Tue, Wed, Thu, Fri-Sun -p "every weekend- dates will be Sat, Sun; periods will be Sat, Sun-Fri day" Depth With the --depth NUM option (short form: -NUM), reports will show ac- counts only to the specified depth, hiding deeper subaccounts. Use this when you want a summary with less detail. This flag has the same effect as a depth: query argument: depth:2, --depth=2 or -2 are equiva- lent. In place of a single number which limits the depth for all accounts, you can also provide separate depth limits for different accounts using regular expressions (since 1.41). For example, --depth assets=2 (or, equivalently: depth:assets=2) will collapse accounts matching the regular expression assets to depth 2. So assets:bank:savings would be collapsed to assets:bank, while liabil- ities:bank:credit card would not be affected. This can be combined with a flat depth to collapse other accounts not matching the regular expression, so --depth assets=2 --depth 1 would collapse as- sets:bank:savings to assets:bank and liabilities:bank:credit card to liabilities. You can supply multiple depth arguments and they will all be applied, so --depth assets=2 --depth liabilities=3 --depth 1 would collapse: o accounts matching assets to depth 2, o accounts matching liabilities to depth 3, o all other accounts to depth 1. If an account is matched by more than one regular expression depth ar- gument then the more specific one will be used. For example, if --depth assets=1 --depth assets:bank:savings=2 is provided, then as- sets:bank:savings will be collapsed to depth 2 rather than depth 1. This is because assets:bank:savings matches at level 3 in the account name, while assets matches at level 1. The same would be true with the argument --depth assets=1 --depth savings=2. Queries Many hledger commands accept query arguments, which restrict their scope and let you report on a precise subset of your data. Here's a quick overview of hledger's queries: o By default, a query argument is treated as a case-insensitive sub- string pattern for matching account names. Eg: dining groceries car:fuel o Patterns containing spaces or other special characters must be en- closed in single or double quotes: 'personal care' o Patterns are actually regular expressions, so you can add regexp metacharacters for more precision (or you may need to backslash-es- cape certain characters; see "Regular expressions" above): '^expenses\b' 'food$' 'fuel|repair' 'accounts (payable|receivable)' o To match something other than the account name, you can add a query type prefix, such as: date:202312- status: desc:amazon cur:USD cur:\\$ amt:'>0' acct:groceries (but acct: is the default, so we usually don't bother writing it) o To negate a query, add a not: prefix: not:status:'*' not:desc:'opening|closing' not:cur:USD o Multiple query terms can be combined, as space-separated queries Eg: hledger print date:2022 desc:amazon desc:amzn (show transactions dated in 2022 whose description contains "amazon" or "amzn"). o Or more flexibly as boolean queries. Eg: hledger print expr:'date:2022 and (desc:amazon or desc:amzn) and not date:202210' All hledger commands use the same query language, but different com- mands may interpret the query in different ways. We haven't described the commands yet (that's coming in PART 4: COMMANDS below) but here's the gist of it: o Transaction-oriented commands (print, aregister, close, import, de- scriptions..) try to match transactions (including the transaction's postings). o Posting-oriented commands (register, balance, balancesheet, incomes- tatement, accounts..) try to match postings. Postings inherit their transaction's attributes for querying purposes, so transaction fields like date or description can still be referenced in a posting query. o A few commands match in more specific ways. (Eg aregister, which has a special first argument.) Query types Here are the query types available: acct: query acct:REGEX, or just REGEX Match account names containing this case insensitive regular expres- sion. This is the default query type, so we usually don't bother writing the "acct:" prefix. amt: query amt:N, amt:'N', amt:'>=N' Match postings with a single-commodity amount equal to, less than, or greater than N. (Postings with multi-commodity amounts are not tested and will always match.) amt: needs quotes to hide the less than/greater than sign from the command line shell. The comparison has two modes: if N is preceded by a + or - sign (or is 0), the two signed numbers are compared. Otherwise, the absolute mag- nitudes are compared, ignoring sign. Keep in mind that amt: matches posting amounts, not account balances. code: query code:REGEX Match by transaction code (eg check number). cur: query cur:REGEX Match postings or transactions including any amounts whose cur- rency/commodity symbol is fully matched by REGEX. (Contrary to hledger's usual infix matching. To do infix matching, write .*REGEX.*.) Note, to match special characters which are regex-signifi- cant, you need to escape them with \. And for characters which are significant to your shell you will usually need one more level of es- caping. Eg to match the dollar sign: cur:\\$ or cur:'\$' desc: query desc:REGEX Match transaction descriptions. date: query date:PERIODEXPR Match dates (or with the --date2 flag, secondary dates) within the specified period. PERIODEXPR is a period expression with no report in- terval. Examples: date:2016, date:thismonth, date:2/1-2/15, date:2021-07-27..nextquarter. date2: query date2:PERIODEXPR If you use secondary dates: this matches secondary dates within the specified period. It is not affected by the --date2 flag. depth: query depth:[REGEXP=]N Match (or display, depending on command) accounts at or above this depth, optionally only for accounts matching a provided regular expres- sion. See Depth for detailed rules. note: query note:REGEX Match transaction notes (the part of the description right of |, or the whole description if there's no |). payee: query payee:REGEX Match transaction payee/payer names (the part of the description left of |, or the whole description if there's no |). real: query real:, real:0 Match real or virtual postings respectively. status: query status:, status:!, status:* Match unmarked, pending, or cleared transactions respectively. type: query type:TYPECODES Match by account type (see Declaring accounts > Account types). TYPE- CODES is one or more of the single-letter account type codes ALERXCV, case insensitive. Note type:A and type:E will also match their respec- tive subtypes C (Cash) and V (Conversion). Certain kinds of account alias can disrupt account types, see Rewriting accounts > Aliases and account types. tag: query tag:NAMEREGEX[=VALREGEX] Match by tag name, and optionally also by tag value. Note: o Both regular expressions do infix matching. If you need a complete match, use ^ and $. Eg: tag:'^fullname$', tag:'^fullname$=^fullvalue$ o To match values, ignoring names, do tag:.=VALREGEX o Accounts also inherit the tags of their parent accounts. o Postings also inherit the tags of their account and their transaction . o Transactions also acquire the tags of their postings. Negative queries not: query not:QUERY You can prepend not: to a query to negate the match. Eg: not:equity, not:desc:apple (Also, a trick: not:not:... can sometimes solve query problems conve- niently.) Space-separated queries When given multiple space-separated query terms, most commands select things which match: o any of the description terms AND o any of the account terms AND o any of the status terms AND o all the other terms. The print command is a little different, showing transactions which: o match any of the description terms AND o have any postings matching any of the positive account terms AND o have no postings matching any of the negative account terms AND o match all the other terms. Boolean queries You can write more complicated "boolean" query expressions, enclosed in quotes and prefixed with expr:. These can combine subqueries with NOT, AND, OR operators (case insensitive), and parentheses for grouping. Eg, to show transactions involving both cash and expense accounts: hledger print expr:'cash AND expenses' The prefix and enclosing quotes are required, so don't write hledger print cash AND expenses. That would be a space-separated query showing transactions involving accounts with any of "cash", "and", "expenses" in their names. You can write space-separated queries inside a boolean query, and they will combine as described above, but it might be confusing and best avoided. Eg these are equivalent, showing transactions involving cash or expenses accounts: hledger print expr:'cash expenses' hledger print cash expenses There is a restriction with date: queries: they may not be used inside OR expressions. Actually, there are three types of boolean query: expr: for general use, and any: and all: variants which can be useful with print. expr: query expr:'QUERYEXPR' For example, expr:'date:lastmonth AND NOT (food OR rent)' means "match things which are dated in the last month and do not have food or rent in the account name". When using expr: with transaction-oriented commands like print, post- ing-oriented query terms like acct: and amt: are considered to match the transaction if they match any of its postings. So, hledger print expr:'cash and amt:>0' means "show transactions with (at least one posting involving a cash account) and (at least one post- ing with a positive amount)". any: query any:'QUERYEXPR' Like expr:, but when used with transaction-oriented commands like print, it matches the transaction only if a posting can be matched by all of QUERYEXPR. So, hledger print any:'cash and amt:>0' means "show transactions where at least one posting posts a positive amount to a cash account". all: query all:'QUERYEXPR' Like expr:, but when used with transaction-oriented commands like print, it matches the transaction only if all postings are matched by all of QUERYEXPR (and there is at least one posting). So, hledger print all:'cash and amt:0' means "show transactions where all postings involve a cash account and have a zero amount". Or, hledger print all:'cash or checking' means "show transactions which touch only cash and/or checking accounts". Queries and command options Some queries can also be expressed as command-line options: depth:2 is equivalent to --depth 2, date:2023 is equivalent to -p 2023, etc. When you mix command options and query arguments, generally the resulting query is their intersection. Queries and account aliases When account names are rewritten with --alias or alias, acct: will match either the old or the new account name. Queries and valuation When amounts are converted to other commodities in cost or value re- ports, cur: and amt: match the old commodity symbol and the old amount quantity, not the new ones. (Except in hledger 1.22, #1625.) Pivoting Normally, hledger groups amounts and displays their totals by account (name). With --pivot PIVOTEXPR, some other field's (or multiple fields') value is used as a synthetic account name, causing different grouping and display. PIVOTEXPR can be o any of these standard transaction or posting fields (their value is substituted): status, code, desc, payee, note, acct, comm/cur, amt, cost o or a tag name o or any combination of these, colon-separated. Some special cases: o Colons appearing in PIVOTEXPR or in a pivoted tag value will generate account hierarchy. o When pivoting a posting has multiple values for a tag, the pivoted value of that tag will be the first value. o When a posting has multiple commodities, the pivoted value of "comm"/"cur" will be "". Also when an unrecognised tag name or field is provided, its pivoted value will be "". (If this causes confusing output, consider excluding those postings from the report.) Examples: 2016/02/16 Yearly Dues Payment assets:bank account 2 EUR income:dues -2 EUR ; member: John Doe, kind: Lifetime Normal balance report showing account names: $ hledger balance 2 EUR assets:bank account -2 EUR income:dues -------------------- 0 Pivoted balance report, using member: tag values instead: $ hledger balance --pivot member 2 EUR -2 EUR John Doe -------------------- 0 One way to show only amounts with a member: value (using a query): $ hledger balance --pivot member tag:member=. -2 EUR John Doe -------------------- -2 EUR Another way (the acct: query matches against the pivoted "account name"): $ hledger balance --pivot member acct:. -2 EUR John Doe -------------------- -2 EUR Hierarchical reports can be generated with multiple pivot values: $ hledger balance Income:Dues --pivot kind:member -2 EUR Lifetime:John Doe -------------------- -2 EUR Generating data hledger can enrich the data provided to it, or generate new data, in a number of ways. Mostly, this is done only if you request it: o Missing amounts or missing costs in transactions are inferred auto- matically when possible. o The --infer-equity flag infers missing conversion equity postings from @/@@ costs. o The --infer-costs flag infers missing costs from conversion equity postings. o The --infer-market-prices flag infers P price directives from costs. o The --auto flag adds extra postings to transactions matched by auto posting rules. o The --forecast option generates transactions from periodic transac- tion rules. o The balance --budget report infers budget goals from periodic trans- action rules. o Commands like close, rewrite, and hledger-interest generate transac- tions or postings. o CSV data is converted to transactions by applying CSV conversion rules.. etc. Such generated data is temporary, existing only at report time. You can convert it to permanent recorded data by, eg, capturing the output of hledger print and saving it in your journal file. This can some- times be useful as a data entry aid. If you are curious what data is being generated and why, run hledger print -x --verbose-tags. -x/--explicit shows inferred amounts and --verbose-tags adds tags like generated-transaction (from periodic rules) and generated-posting, modified (from auto posting rules). Sim- ilar hidden tags (with an underscore prefix) are always present, also, so you can always match such data with queries like tag:generated or tag:modified. Forecasting Forecasting, or speculative future reporting, can be useful for esti- mating future balances, or for exploring different future scenarios. The simplest and most flexible way to do it with hledger is to manually record a bunch of future-dated transactions. You could keep these in a separate future.journal and include that with -f only when you want to see them. --forecast There is another way: with the --forecast option, hledger can generate temporary "forecast transactions" for reporting purposes, according to periodic transaction rules defined in the journal. Each rule can gen- erate multiple recurring transactions, so by changing one rule you can change many forecasted transactions. Forecast transactions usually start after ordinary transactions end. By default, they begin after your latest-dated ordinary transaction, or today, whichever is later, and they end six months from today. (The exact rules are a little more complicated, and are given below.) This is the "forecast period", which need not be the same as the report period. You can override it - eg to forecast farther into the future, or to force forecast transactions to overlap your ordinary transactions - by giving the --forecast option a period expression argument, like --forecast=..2099 or --forecast=2023-02-15... Note that the = is re- quired. Inspecting forecast transactions print is the best command for inspecting and troubleshooting forecast transactions. Eg: ~ monthly from 2022-12-20 rent assets:bank:checking expenses:rent $1000 $ hledger print --forecast --today=2023/4/21 2023-05-20 rent ; generated-transaction: ~ monthly from 2022-12-20 assets:bank:checking expenses:rent $1000 2023-06-20 rent ; generated-transaction: ~ monthly from 2022-12-20 assets:bank:checking expenses:rent $1000 2023-07-20 rent ; generated-transaction: ~ monthly from 2022-12-20 assets:bank:checking expenses:rent $1000 2023-08-20 rent ; generated-transaction: ~ monthly from 2022-12-20 assets:bank:checking expenses:rent $1000 2023-09-20 rent ; generated-transaction: ~ monthly from 2022-12-20 assets:bank:checking expenses:rent $1000 Here there are no ordinary transactions, so the forecasted transactions begin on the first occurrence after today's date. (You won't normally use --today; it's just to make these examples reproducible.) Forecast reports Forecast transactions affect all reports, as you would expect. Eg: $ hledger areg rent --forecast --today=2023/4/21 Transactions in expenses:rent and subaccounts: 2023-05-20 rent as:ba:checking $1000 $1000 2023-06-20 rent as:ba:checking $1000 $2000 2023-07-20 rent as:ba:checking $1000 $3000 2023-08-20 rent as:ba:checking $1000 $4000 2023-09-20 rent as:ba:checking $1000 $5000 $ hledger bal -M expenses --forecast --today=2023/4/21 Balance changes in 2023-05-01..2023-09-30: || May Jun Jul Aug Sep ===============++=================================== expenses:rent || $1000 $1000 $1000 $1000 $1000 ---------------++----------------------------------- || $1000 $1000 $1000 $1000 $1000 Forecast tags Forecast transactions generated by --forecast have a hidden tag, _gen- erated-transaction. So if you ever need to match forecast transac- tions, you could use tag:_generated-transaction (or just tag:generated) in a query. For troubleshooting, you can add the --verbose-tags flag. Then, visi- ble generated-transaction tags will be added also, so you can view them with the print command. Their value indicates which periodic rule was responsible. Forecast period, in detail Forecast start/end dates are chosen so as to do something useful by de- fault in almost all situations, while also being flexible. Here are (with luck) the exact rules, to help with troubleshooting: The forecast period starts on: o the later of o the start date in the periodic transaction rule o the start date in --forecast's argument o otherwise (if those are not available): the later of o the report start date specified with -b/-p/date: o the day after the latest ordinary transaction in the journal o otherwise (if none of these are available): today. The forecast period ends on: o the earlier of o the end date in the periodic transaction rule o the end date in --forecast's argument o otherwise: the report end date specified with -e/-p/date: o otherwise: 180 days (~6 months) from today. Forecast troubleshooting When --forecast is not doing what you expect, one of these tips should help: o Remember to use the --forecast option. o Remember to have at least one periodic transaction rule in your jour- nal. o Test with print --forecast. o Check for typos or too-restrictive start/end dates in your periodic transaction rule. o Leave at least 2 spaces between the rule's period expression and de- scription fields. o Check for future-dated ordinary transactions suppressing forecasted transactions. o Try setting explicit report start and/or end dates with -b, -e, -p or date: o Try adding the -E flag to encourage display of empty periods/zero transactions. o Try setting explicit forecast start and/or end dates with --fore- cast=START..END o Consult Forecast period, in detail, above. o Check inside the engine: add --debug=2 (eg). Budgeting With the balance command's --budget report, each periodic transaction rule generates recurring budget goals in specified accounts, and goals and actual performance can be compared. See the balance command's doc below. You can generate budget goals and forecast transactions at the same time, from the same or different periodic transaction rules: hledger bal -M --budget --forecast ... See also: Budgeting and Forecasting. Amount formatting Commodity display style For the amounts in each commodity, hledger chooses a consistent display style (symbol placement, decimal mark and digit group marks, number of decimal digits) to use in most reports. This is inferred as follows: First, if there's a D directive declaring a default commodity, that commodity symbol and amount format is applied to all no-symbol amounts in the journal. Then each commodity's display style is determined from its commodity directive. We recommend always declaring commodities with commodity directives, since they help ensure consistent display styles and preci- sions, and bring other benefits such as error checking for commodity symbols. Here's an example: # Set display styles (and decimal marks, for parsing, if there is no decimal-mark directive) # for the $, EUR, INR and no-symbol commodities: commodity $1,000.00 commodity EUR 1.000,00 commodity INR 9,99,99,999.00 commodity 1 000 000.9455 But for convenience, if a commodity directive is not present, hledger infers a commodity's display styles from its amounts as they are writ- ten in the journal (excluding cost amounts and amounts in periodic transaction rules or auto posting rules). It uses o the symbol placement and decimal mark of the first amount seen o the digit group marks of the first amount with digit group marks o and the maximum number of decimal digits seen across all amounts. And as fallback if no applicable amounts are found, it would use a de- fault style, like $1000.00 (symbol on the left with no space, period as decimal mark, and two decimal digits). Finally, commodity styles can be overridden by the -c/--commodity-style command line option. Rounding Amounts are stored internally as decimal numbers with up to 255 decimal places. They are displayed with their original journal precisions by print and print-like reports, and rounded to their display precision (the number of decimal digits specified by the commodity display style) by other reports. When rounding, hledger uses banker's rounding (it rounds to the nearest even digit). So eg 0.5 displayed with zero deci- mal digits appears as "0". Trailing decimal marks If you're wondering why your print report sometimes shows trailing dec- imal marks, with no decimal digits; it does this when showing amounts that have digit group marks but no decimal digits, to disambiguate them and allow them to be re-parsed reliably (see Decimal marks). Eg: commodity $1,000.00 2023-01-02 (a) $1000 $ hledger print 2023-01-02 (a) $1,000. If this is a problem (eg when exporting to Ledger), you can avoid it by disabling digit group marks, eg with -c/--commodity (for each affected commodity): $ hledger print -c '$1000.00' 2023-01-02 (a) $1000 or by forcing print to always show decimal digits, with --round: $ hledger print -c '$1,000.00' --round=soft 2023-01-02 (a) $1,000.00 Amount parseability More generally, hledger output falls into three rough categories, which format amounts a little bit differently to suit different consumers: 1. "hledger-readable output" - should be readable by hledger (and by humans) o This is produced by reports that show full journal entries: print, import, close, rewrite etc. o It shows amounts with their original journal precisions, which may not be consistent from one amount to the next. o It adds a trailing decimal mark when needed to avoid showing ambigu- ous amounts. o It can be parsed reliably (by hledger and ledger2beancount at least, but perhaps not by Ledger..) 2. "human-readable output" - usually for humans o This is produced by all other reports. o It shows amounts with standard display precisions, which will be con- sistent within each commodity. o It shows ambiguous amounts unmodified. o It can be parsed reliably in the context of a known report (when you know decimals are consistently not being shown, you can assume a sin- gle mark is a digit group mark). 3. "machine-readable output" - usually for other software o This is produced by all reports when an output format like csv, tsv, json, or sql is selected. o It shows amounts as 1 or 2 do, but without digit group marks. o It can be parsed reliably (if needed, the decimal mark can be changed with -c/--commodity-style). Cost reporting In some transactions - for example a currency conversion, or a purchase or sale of stock - one commodity is exchanged for another. In these transactions there is a conversion rate, also called the cost (when buying) or selling price (when selling). (In hledger docs we just say "cost" generically for convenience.) With the -B/--cost flag, hledger can show amounts "at cost", converted to the cost's commodity. Recording costs We'll explore several ways of recording transactions involving costs. These are also summarised at hledger Cookbook > Cost notation. Costs can be recorded explicitly in the journal, using the @ UNITCOST or @@ TOTALCOST notation described in Journal > Costs: Variant 1 2022-01-01 assets:dollars $-135 assets:euros 100 @ $1.35 ; $1.35 per euro (unit cost) Variant 2 2022-01-01 assets:dollars $-135 assets:euros 100 @@ $135 ; $135 total cost Typically, writing the unit cost (variant 1) is preferable; it can be more effort, requiring more attention to decimal digits; but it reveals the per-unit cost basis, and makes stock sales easier. Costs can also be left implicit, and hledger will infer the cost that is consistent with a balanced transaction: Variant 3 2022-01-01 assets:dollars $-135 assets:euros 100 Here, hledger will attach a @@ 100 cost to the first amount (you can see it with hledger print -x). This form looks convenient, but there are downsides: o It sacrifices some error checking. For example, if you accidentally wrote 10 instead of 100, hledger would not be able to detect the mis- take. o It is sensitive to the order of postings - if they were reversed, a different entry would be inferred and reports would be different. o The per-unit cost basis is not easy to read. So generally this kind of entry is not recommended. You can make sure you have none of these by using -s (strict mode), or by running hledger check balanced. Reporting at cost Now when you add the -B/--cost flag to reports ("B" is from Ledger's -B/--basis/--cost flag), any amounts which have been annotated with costs will be converted to their cost's commodity (in the report out- put). Ie they will be displayed "at cost" or "at sale price". Some things to note: o Costs are attached to specific posting amounts in specific transac- tions, and once recorded they do not change. This contrasts with market prices, which are ambient and fluctuating. o Conversion to cost is performed before conversion to market value (described below). Equity conversion postings There is a problem with the entries above - they are not conventional Double Entry Bookkeeping (DEB) notation, and because of the "magical" transformation of one commodity into another, they cause an imbalance in the Accounting Equation. This shows up as a non-zero grand total in balance reports like hledger bse. For most hledger users, this doesn't matter in practice and can safely be ignored ! But if you'd like to learn more, keep reading. Conventional DEB uses an extra pair of equity postings to balance the transaction. Of course you can do this in hledger as well: Variant 4 2022-01-01 assets:dollars $-135 assets:euros 100 equity:conversion $135 equity:conversion -100 Now the transaction is perfectly balanced according to standard DEB, and hledger bse's total will not be disrupted. And, hledger can still infer the cost for cost reporting, but it's not done by default - you must add the --infer-costs flag like so: $ hledger print --infer-costs 2022-01-01 one hundred euros purchased at $1.35 each assets:dollars $-135 @@ 100 assets:euros 100 equity:conversion $135 equity:conversion -100 $ hledger bal --infer-costs -B -100 assets:dollars 100 assets:euros -------------------- 0 Here are some downsides of this kind of entry: o The per-unit cost basis is not easy to read. o Instead of -B you must remember to type -B --infer-costs. o --infer-costs works only where hledger can identify the two eq- uity:conversion postings and match them up with the two non-equity postings. So writing the journal entry in a particular format be- comes more important. More on this below. Inferring equity conversion postings Can we go in the other direction ? Yes, if you have transactions writ- ten with the @/@@ cost notation, hledger can infer the missing equity postings, if you add the --infer-equity flag. Eg: 2022-01-01 assets:dollars -$135 assets:euros 100 @ $1.35 $ hledger print --infer-equity 2022-01-01 assets:dollars $-135 assets:euros 100 @ $1.35 equity:conversion:$-: -100 equity:conversion:$-:$ $135.00 The equity account names will be "equity:conversion:A-B:A" and "eq- uity:conversion:A-B:B" where A is the alphabetically first commodity symbol. You can customise the "equity:conversion" part by declaring an account with the V/Conversion account type. Note you will need to add account declarations for these to your jour- nal, if you use check accounts or check --strict. Combining costs and equity conversion postings Finally, you can use both the @/@@ cost notation and equity postings at the same time. This in theory gives the best of all worlds - preserv- ing the accounting equation, revealing the per-unit cost basis, and providing more flexibility in how you write the entry: Variant 5 2022-01-01 one hundred euros purchased at $1.35 each assets:dollars $-135 equity:conversion $135 equity:conversion -100 assets:euros 100 @ $1.35 All the other variants above can (usually) be rewritten to this final form with: $ hledger print -x --infer-costs --infer-equity Downsides: o The precise format of the journal entry becomes more important. If hledger can't detect and match up the cost and equity postings, it will give a transaction balancing error. o The add command does not yet accept this kind of entry (#2056). o This is the most verbose form. Requirements for detecting equity conversion postings --infer-costs has certain requirements (unlike --infer-equity, which always works). It will infer costs only in transactions with: o Two non-equity postings, in different commodities. Their order is significant: the cost will be added to the first of them. o Two postings to equity conversion accounts, next to one another, which balance the two non-equity postings. This balancing is checked to the same precision (number of decimal places) used in the conver- sion posting's amount. Equity conversion accounts are: o any accounts declared with account type V/Conversion, or their sub- accounts o otherwise, accounts named equity:conversion, equity:trade, or eq- uity:trading, or their subaccounts. And multiple such four-posting groups can coexist within a single transaction. When --infer-costs fails, it does not infer a cost in that transaction, and does not raise an error (ie, it infers costs where it can). Reading variant 5 journal entries, combining cost notation and equity postings, has all the same requirements. When reading such an entry fails, hledger raises an "unbalanced transaction" error. Infer cost and equity by default ? Should --infer-costs and --infer-equity be enabled by default ? Try using them always, eg with a shell alias: alias h="hledger --infer-equity --infer-costs" and let us know what problems you find. Value reporting hledger can also show amounts "at market value", converted to some other commodity using the market price or conversion rate on a certain date. This is controlled by the --value=TYPE[,COMMODITY] option. We also provide simpler -V and -X COMMODITY aliases for this, which are often sufficient. The market prices are declared with a special P directive, and/or they can be inferred from the costs recorded in transactions, by using the --infer-market-prices flag. -V: Value The -V/--market flag converts amounts to market value in their default valuation commodity, using the market prices in effect on the valuation date(s), if any. More on these in a minute. -X: Value in specified commodity The -X/--exchange=COMM option is like -V, except you tell it which cur- rency you want to convert to, and it tries to convert everything to that. Valuation date Market prices can change from day to day. hledger will use the prices on a particular valuation date (or on more than one date). By default hledger uses "end" dates for valuation. More specifically: o For single period reports (including normal print and register re- ports): o If an explicit report end date is specified, that is used. o Otherwise the latest transaction date or non-future P directive date is used. o For multiperiod reports, each period is valued on its last day. This can be customised with the --value option described below, which can select either "then", "end", "now", or "custom" dates. Finding market price To convert a commodity A to its market value in another commodity B, hledger looks for a suitable market price (exchange rate) as follows, in this order of preference: 1. A declared market price or inferred market price: A's latest market price in B on or before the valuation date as declared by a P direc- tive, or (with the --infer-market-prices flag) inferred from costs. 2. A reverse market price: the inverse of a declared or inferred market price from B to A. 3. A forward chain of market prices: a synthetic price formed by com- bining the shortest chain of "forward" (only 1 above) market prices, leading from A to B. 4. Any chain of market prices: a chain of any market prices, including both forward and reverse prices (1 and 2 above), leading from A to B. There is a limit to the length of these price chains; if hledger reaches that length without finding a complete chain or exhausting all possibilities, it will give up (with a "gave up" message visible in --debug=2 output). That limit is currently 1000. Amounts for which no suitable market price can be found, are not con- verted. --infer-market-prices: market prices from transactions Normally, market value in hledger is fully controlled by, and requires, P directives in your journal. Since adding and updating those can be a chore, and since transactions usually take place at close to market value, why not use the recorded costs as additional market prices (as Ledger does) ? Adding the --infer-market-prices flag to -V, -X or --value enables this. So for example, hledger bs -V --infer-market-prices will get market prices both from P directives and from transactions. If both occur on the same day, the P directive takes precedence. There is a downside: value reports can sometimes be affected in confus- ing/undesired ways by your journal entries. If this happens to you, read all of this Value reporting section carefully, and try adding --debug or --debug=2 to troubleshoot. --infer-market-prices can infer market prices from: o multicommodity transactions with explicit prices (@/@@) o multicommodity transactions with implicit prices (no @, two commodi- ties, unbalanced). (With these, the order of postings matters. hledger print -x can be useful for troubleshooting.) o multicommodity transactions with equity postings, if cost is inferred with --infer-costs. There is a limitation (bug) currently: when a valuation commodity is not specified, prices inferred with --infer-market-prices do not help select a default valuation commodity, as P prices would. So conversion might not happen because no valuation commodity was detected (--debug=2 will show this). To be safe, specify the valuation commmodity, eg: o -X EUR --infer-market-prices, not -V --infer-market-prices o --value=then,EUR --infer-market-prices, not --value=then --infer-mar- ket-prices Signed costs and market prices can be confusing. For reference, here is the current behaviour, since hledger 1.25. (If you think it should work differently, see #1870.) 2022-01-01 Positive Unit prices a A 1 b B -1 @ A 1 2022-01-01 Positive Total prices a A 1 b B -1 @@ A 1 2022-01-02 Negative unit prices a A 1 b B 1 @ A -1 2022-01-02 Negative total prices a A 1 b B 1 @@ A -1 2022-01-03 Double Negative unit prices a A -1 b B -1 @ A -1 2022-01-03 Double Negative total prices a A -1 b B -1 @@ A -1 All of the transactions above are considered balanced (and on each day, the two transactions are considered equivalent). Here are the market prices inferred for B: $ hledger -f- --infer-market-prices prices P 2022-01-01 B A 1 P 2022-01-01 B A 1.0 P 2022-01-02 B A -1 P 2022-01-02 B A -1.0 P 2022-01-03 B A -1 P 2022-01-03 B A -1.0 Valuation commodity When you specify a valuation commodity (-X COMM or --value TYPE,COMM): hledger will convert all amounts to COMM, wherever it can find a suit- able market price (including by reversing or chaining prices). When you leave the valuation commodity unspecified (-V or --value TYPE): For each commodity A, hledger picks a default valuation commodity as follows, in this order of preference: 1. The price commodity from the latest P-declared market price for A on or before valuation date. 2. The price commodity from the latest P-declared market price for A on any date. (Allows conversion to proceed when there are inferred prices before the valuation date.) 3. If there are no P directives at all (any commodity or date) and the --infer-market-prices flag is used: the price commodity from the latest transaction-inferred price for A on or before valuation date. This means: o If you have P directives, they determine which commodities -V will convert, and to what. o If you have no P directives, and use the --infer-market-prices flag, costs determine it. Amounts for which no valuation commodity can be found are not con- verted. --value: Flexible valuation -V and -X are special cases of the more general --value option: --value=TYPE[,COMM] TYPE is then, end, now or YYYY-MM-DD. COMM is an optional commodity symbol. Shows amounts converted to: - default valuation commodity (or COMM) using market prices at posting dates - default valuation commodity (or COMM) using market prices at period end(s) - default valuation commodity (or COMM) using current market prices - default valuation commodity (or COMM) using market prices at some date The TYPE part selects cost or value and valuation date: --value=then Convert amounts to their value in the default valuation commod- ity, using market prices on each posting's date. --value=end Convert amounts to their value in the default valuation commod- ity, using market prices on the last day of the report period (or if unspecified, the journal's end date); or in multiperiod reports, market prices on the last day of each subperiod. --value=now Convert amounts to their value in the default valuation commod- ity using current market prices (as of when report is gener- ated). --value=YYYY-MM-DD Convert amounts to their value in the default valuation commod- ity using market prices on this date. To select a different valuation commodity, add the optional ,COMM part: a comma, then the target commodity's symbol. Eg: --value=now,EUR. hledger will do its best to convert amounts to this commodity, deducing market prices as described above. Valuation examples Here are some quick examples of -V: ; one euro is worth this many dollars from nov 1 P 2016/11/01 $1.10 ; purchase some euros on nov 3 2016/11/3 assets:euros 100 assets:checking ; the euro is worth fewer dollars by dec 21 P 2016/12/21 $1.03 How many euros do I have ? $ hledger -f t.j bal -N euros 100 assets:euros What are they worth at end of nov 3 ? $ hledger -f t.j bal -N euros -V -e 2016/11/4 $110.00 assets:euros What are they worth after 2016/12/21 ? (no report end date specified, defaults to today) $ hledger -f t.j bal -N euros -V $103.00 assets:euros Here are some examples showing the effect of --value, as seen with print: P 2000-01-01 A 1 B P 2000-02-01 A 2 B P 2000-03-01 A 3 B P 2000-04-01 A 4 B 2000-01-01 (a) 1 A @ 5 B 2000-02-01 (a) 1 A @ 6 B 2000-03-01 (a) 1 A @ 7 B Show the cost of each posting: $ hledger -f- print --cost 2000-01-01 (a) 5 B 2000-02-01 (a) 6 B 2000-03-01 (a) 7 B Show the value as of the last day of the report period (2000-02-29): $ hledger -f- print --value=end date:2000/01-2000/03 2000-01-01 (a) 2 B 2000-02-01 (a) 2 B With no report period specified, the latest transaction date or price date is used as valuation date (2000-04-01): $ hledger -f- print --value=end 2000-01-01 (a) 3 B 2000-02-01 (a) 3 B 2000-03-01 (a) 3 B The value today is the same (the 2000-04-01 price is still in effect): $ hledger -f- print --value=now 2000-01-01 (a) 4 B 2000-02-01 (a) 4 B 2000-03-01 (a) 4 B Show the value on 2000/01/15: $ hledger -f- print --value=2000-01-15 2000-01-01 (a) 1 B 2000-02-01 (a) 1 B 2000-03-01 (a) 1 B Interaction of valuation and queries When matching postings based on queries in the presence of valuation, the following happens: 1. The query is separated into two parts: 1. the currency (cur:) or amount (amt:). 2. all other parts. 2. The postings are matched to the currency and amount queries based on pre-valued amounts. 3. Valuation is applied to the postings. 4. The postings are matched to the other parts of the query based on post-valued amounts. Related: #1625 Effect of valuation on reports Here is a reference for how valuation is supposed to affect each part of hledger's reports. (It's wide, you may need to scroll sideways.) It may be useful when troubleshooting. If you find problems, please report them, ideally with a reproducible example. Related: #329, #1083. First, a quick glossary: cost calculated using price(s) recorded in the transaction(s). value market value using available market price declarations, or the unchanged amount if no conversion rate can be found. report start the first day of the report period specified with -b or -p or date:, otherwise today. report or journal start the first day of the report period specified with -b or -p or date:, otherwise the earliest transaction date in the journal, otherwise today. report end the last day of the report period specified with -e or -p or date:, otherwise today. report or journal end the last day of the report period specified with -e or -p or date:, otherwise the latest transaction date in the journal, otherwise today. report interval a flag (-D/-W/-M/-Q/-Y) or period expression that activates the report's multi-period mode (whether showing one or many subperi- ods). Report -B, --cost -V, -X --value=then --value=end --value=DATE, type --value=now -------------------------------------------------------------------------------------------- print posting cost value at re- value at posting value at re- value at amounts port end or date port or DATE/today today journal end balance unchanged unchanged unchanged unchanged unchanged asser- tions/as- signments register starting cost value at re- valued at day value at re- value at balance port or each historical port or DATE/today (-H) journal end posting was made journal end starting cost value at day valued at day value at day value at balance before re- each historical before re- DATE/today (-H) with port or posting was made port or report journal journal interval start start posting cost value at re- value at posting value at re- value at amounts port or date port or DATE/today journal end journal end summary summarised value at pe- sum of postings value at pe- value at posting cost riod ends in interval, val- riod ends DATE/today amounts ued at interval with re- start port in- terval running sum/average sum/average sum/average of sum/average sum/average total/av- of displayed of displayed displayed values of displayed of displayed erage values values values values balance (bs, bse, cf, is) balance sums of value at re- value at posting value at re- value at changes costs port end or date port or DATE/today of today of journal end sums of post- sums of of sums of ings postings postings budget like balance like balance like balance like bal- like balance amounts changes changes changes ances changes (--bud- get) grand to- sum of dis- sum of dis- sum of displayed sum of dis- sum of dis- tal played val- played val- valued played val- played values ues ues ues balance (bs, bse, cf, is) with re- port in- terval starting sums of value at re- sums of values of value at re- sums of post- balances costs of port start postings before port start ings before (-H) postings be- of sums of report start at of sums of report start fore report all postings respective post- all postings start before re- ing dates before re- port start port start balance sums of same as sums of values of balance value at changes costs of --value=end postings in pe- change in DATE/today of (bal, is, postings in riod at respec- each period, sums of post- bs period tive posting valued at ings --change, dates period ends cf --change) end bal- sums of same as sums of values of period end value at ances costs of --value=end postings from be- balances, DATE/today of (bal -H, postings fore period start valued at sums of post- is --H, from before to period end at period ends ings bs, cf) report start respective post- to period ing dates end budget like balance like balance like balance like bal- like balance amounts changes/end changes/end changes/end bal- ances changes/end (--bud- balances balances ances balances get) row to- sums, aver- sums, aver- sums, averages of sums, aver- sums, aver- tals, row ages of dis- ages of dis- displayed values ages of dis- ages of dis- averages played val- played val- played val- played values (-T, -A) ues ues ues column sums of dis- sums of dis- sums of displayed sums of dis- sums of dis- totals played val- played val- values played val- played values ues ues ues grand to- sum, average sum, average sum, average of sum, average sum, average tal, of column of column column totals of column of column to- grand av- totals totals totals tals erage --cumulative is omitted to save space, it works like -H but with a zero starting balance. PART 4: COMMANDS Here are hledger's standard subcommands. You can list these by running hledger. If you have installed more add-on commands, they also will be listed. In the following command docs, each command's specific options are shown. Most commands also support the general options described above, though some of them might have no effect. (Usually if there's a sensi- ble way for a general option to affect a command, it will.) You can list all of a command's options by running hledger CMD -h. Help commands o commands - show the hledger commands list (default) o demo - show small hledger demos in the terminal o help - show the hledger manual with info, man, or pager User interface commands o repl - run commands from an interactive prompt o run - run commands from a script o ui - (if installed) run hledger's terminal UI o web - (if installed) run hledger's web UI Data entry commands o add - add transactions using terminal prompts o import - add new transactions from other files, eg CSV files Basic report commands o accounts - show account names o codes - show transaction codes o commodities - show commodity/currency symbols o descriptions - show transaction descriptions o files - show input file paths o notes - show note parts of transaction descriptions o payees - show payee parts of transaction descriptions o prices - show market prices o stats - show journal statistics o tags - show tag names Standard report commands o print - show transactions or export journal data o aregister (areg) - show transactions in a particular account o register (reg) - show postings in one or more accounts & running to- tal o balancesheet (bs) - show assets, liabilities and net worth o balancesheetequity (bse) - show assets, liabilities and equity o cashflow (cf) - show changes in liquid assets o incomestatement (is) - show revenues and expenses Advanced report commands o balance (bal) - show balance changes, end balances, budgets, gains.. o roi - show return on investments Chart commands o activity - show bar charts of posting counts per period Data generation commands o close - generate balance-zeroing/restoring transactions o rewrite - generate auto postings, like print --auto Maintenance commands o check - check for various kinds of error in the data o diff - compare account transactions in two journal files o setup - check and show the status of the hledger installation o test - run self tests Next, these commands are described in detail. Help commands commands Show the hledger commands list. Flags: --builtin show only builtin commands, not addons demo Play demos of hledger usage in the terminal, if asciinema is installed. Flags: -s --speed=SPEED playback speed (1 is original speed, .5 is half, 2 is double, etc (default: 2)) Run this command with no argument to list the demos. To play a demo, write its number or a prefix or substring of its title. Tips: Make your terminal window large enough to see the demo clearly. Use the -s/--speed SPEED option to set your preferred playback speed, eg -s4 to play at 4x original speed or -s.5 to play at half speed. The default speed is 2x. During playback, several keys are available: SPACE to pause/unpause, . to step forward (while paused), CTRL-c quit. Examples: $ hledger demo # list available demos $ hledger demo 1 # play the first demo at default speed (2x) $ hledger demo install -s4 # play the "install" demo at 4x speed This command is experimental: there aren't many useful demos yet. help Show the hledger user manual with info, man, or a pager. With a (case insensitive) TOPIC argument, try to open it at that section heading. Flags: -i show the manual with info -m show the manual with man -p show the manual with $PAGER or less (less is always used if TOPIC is specified) This command shows the hledger manual built in to your hledger exe- cutable. It can be useful when offline, or when you prefer the termi- nal to a web browser, or when the appropriate hledger manual or viewers are not installed properly on your system. By default it chooses the best viewer found in $PATH, trying in this order: info, man, $PAGER, less, more, stdout. (If a TOPIC is speci- fied, $PAGER and more are not tried.) You can force the use of info, man, or a pager with the -i, -m, or -p flags. If no viewer can be found, or if running non-interactively, it just prints the manual to stdout. When using info, TOPIC can match either the full heading or a prefix. If your info --version is < 6, you'll need to upgrade it, eg with 'brew install texinfo' on mac. When using man or less, TOPIC must match the full heading. For a pre- fix match, you can write 'TOPIC.*'. Examples $ hledger help -h # show the help command's usage $ hledger help # show the manual with info, man or $PAGER $ hledger help 'time periods' # show the manual's "Time periods" topic $ hledger help 'time periods' -m # use man, even if info is installed User interface commands repl Start an interactive prompt, where you can run any of hledger's com- mands. Data files are parsed just once, so the commands run faster. Flags: no command-specific flags This command is experimental and could change in the future. hledger repl starts a read-eval-print loop (REPL) where you can enter commands interactively. As with the run command, each input file (or each input file/input options combination) is parsed just once, so com- mands will run more quickly than if you ran them individually at the command line. Also like run, the input file(s) specified for the repl command will be the default input for all interactive commands. You can override this temporarily by specifying an -f option in particular commands. But note that commands will not see any changes made to input files (eg by add) until you exit and restart the REPL. The command syntax is the same as with run: o enter one hledger command at a time, without the usual hledger first word o empty lines and comment text from # to end of line are ignored o use single or double quotes to quote arguments when needed o type exit or quit or control-D to exit the REPL. While it is running, the REPL remembers your command history, and you can navigate in the usual ways: o Keypad or Emacs navigation keys to edit the current command line o UP/DOWN or control-P/control-N to step back/forward through history o control-R to search for a past command o TAB to complete file paths. Generally repl command lines should feel much like the normal hledger CLI, but you may find differences. repl is a little stricter; eg it requires full command names or official abbreviations (as seen in the commands list). The commands and help commands, and the command help flags (CMD --tldr, CMD -h/--help, CMD --info, CMD --man), can be useful. You can type control-C to cancel a long-running command (but only once; typing it a second time will exit the REPL). And in most shells you can type control-Z to temporarily exit to the shell (and then fg to return to the REPL). Examples Start the REPL and enter some commands: $ hledger repl Enter hledger commands. To exit, enter 'quit' or 'exit', or send EOF. % stats Main file : .../2025.journal ... % stats -f 2024/2024.journal Main file : .../2024.journal ... % stats Main file : .../2025.journal ... or: $ hledger repl -f some.journal Enter hledger commands. To exit, enter 'quit' or 'exit', or send EOF. % bs ... % print -b 'last week' ... % bs -f other.journal ... run Run a sequence of hledger commands, provided as files or command line arguments. Data files are parsed just once, so the commands run faster. Flags: no command-specific flags This command is experimental and could change in the future. You can use run in three ways: o hledger run -- CMD1 -- CMD2 -- CMD3 - read commands from the command line, separated by -- o hledger run SCRIPTFILE1 SCRIPTFILE2 - read commands from one or more files o cat SCRIPTFILE1 | hledger run - read commands from standard input. run first loads the input file(s) specified by LEDGER_FILE or by -f op- tions, in the usual way. Then it runs each command in turn, each using the same input data. But if you want a particular command to use dif- ferent input, you can specify an -f option within that command. This will override (not add to) the default input, just for that command. Each input file (more precisely, each combination of input file and in- put options) is parsed only once. This means that commands will not see any changes made to these files, until the next run. But the com- mands will run more quickly than if run individually (typically about twice as fast). Command scripts, whether in a file or written on the command line, have a simple syntax: o each line may contain a single hledger command and its arguments, without the usual hledger first word o empty lines are ignored o text from # to end of line is a comment, and ignored o you can use single or double quotes to quote arguments when needed, as on the command line o these extra commands are available: echo TEXT prints some text, and exit or quit ends the run. On unix systems you can use #!/usr/bin/env hledger run in the first line of a command file to make it a runnable script. If that gives an error, use #!/usr/bin/env -S hledger run. It's ok to use the run command recursively within a command script. You may find some differences in behaviour between run command lines and normal hledger command lines. run is a little stricter; eg it re- quires full command names or official abbreviations (as seen in the commands list), and command options must be written after the command name. Examples Run commands from the command line: hledger -f some.journal run -- balance assets --depth 2 -- balance liabilities -f /some/other.journal --depth 3 --transpose -- stats This would load some.journal, run balance assets --depth 2 on it, then run balance liabilities --depth 3 --transpose on /some/other.journal, and finally run stats on some.journal Run commands from standard input: (echo "files"; echo "stats") | hledger -f some.journal run Run commands as a script: $ cat report #!/usr/bin/env -S hledger run -f some.journal echo "List of accounts in some.journal" accounts echo "Assets of some.journal" balance assets --depth 2 echo "Liabilities from /some/other.journal" balance liabilities -f /some/other.journal --depth 3 --transpose echo "Commands from another.script, applied to another.journal" run -f another.journal another.script $ chmod +x report $ ./report List of accounts in some.journal ... ui Runs hledger-ui (if installed). web Runs hledger-web (if installed). Data entry commands add Add new transactions to a journal file, with interactive prompting. Flags: --no-new-accounts don't allow creating new accounts Many hledger users edit their journals directly with a text editor, or generate them from CSV. For more interactive data entry, there is the add command, which prompts interactively on the console for new trans- actions, and appends them to the main journal file (which should be in journal format). Existing transactions are not changed. This is one of the few hledger commands that writes to the journal file (see also import). To use it, just run hledger add and follow the prompts. You can add as many transactions as you like; when you are finished, enter . or press control-d or control-c to exit. Features: o add tries to provide useful defaults, using the most similar (by de- scription) recent transaction (filtered by the query, if any) as a template. o You can also set the initial defaults with command line arguments. o Readline-style edit keys can be used during data entry. o The tab key will auto-complete whenever possible - accounts, pay- ees/descriptions, dates (yesterday, today, tomorrow). If the input area is empty, it will insert the default value. o A parenthesised transaction code may be entered following a date. o Comments and tags may be entered following a description or amount. o If you make a mistake, enter < at any prompt to go one step backward. o Input prompts are displayed in a different colour when the terminal supports it. Notes: o If you enter a number with no commodity symbol, and you have declared a default commodity with a D directive, you might expect add to add this symbol for you. It does not do this; we assume that if you are using a D directive you prefer not to see the commodity symbol re- peated on amounts in the journal. o add creates entries in journal format; it won't work with timeclock or timedot files. Examples: o Record new transactions, saving to the default journal file: hledger add o Add transactions to 2024.journal, but also load 2023.journal for com- pletions: hledger add --file 2024.journal --file 2023.journal o Provide answers for the first four prompts: hledger add today 'best buy' expenses:supplies '$20' There is a detailed tutorial at https://hledger.org/add.html. add and balance assertions Since hledger 1.43, whenever you enter a posting amount, add will re-check all balance assertions in the journal, and if any of them fail, it will report the problem and ask for the amount again. You can also add a new balance assertion, following the amount as in journal format. The new transaction's date, and the new posting's posting date if any (entered in a comment following the amount), will influence assertion checking. You can use -I/--ignore-assertions to disable assertion checking tem- porarily. add and balance assignments Balance assignments are not recalculated during a hledger add session. When add runs, it sees the journal with all balance assignments already processed and converted to assertions. So if you add a new posting which is dated earlier than a balance assignment, it will break the as- sertion and be rejected. You can make it work by using hledger add -I. import Import new transactions from one or more data files to the main jour- nal. Flags: --catchup just mark all transactions as already imported --dry-run just show the transactions to be imported This command detects new transactions in one or more data files speci- fied as arguments, and appends them to the main journal. You can import from any input file format hledger supports, but CSV/SSV/TSV files, downloaded from financial institutions, are the most common import source. The import destination is the default journal file, or another speci- fied in the usual way with $LEDGER_FILE or -f/--file. It should be in journal format. Examples: $ hledger import bank1-checking.csv bank1-savings.csv $ hledger import *.csv Import dry run It's useful to preview the import by running first with --dry-run, to sanity check the range of dates being imported, and to check the effect of your conversion rules if converting from CSV. Eg: $ hledger import bank.csv --dry-run The dry run output is valid journal format, so hledger can re-parse it. If the output is large, you could show just the uncategorised transac- tions like so: $ hledger import --dry-run bank.csv | hledger -f- -I print unknown You could also run this repeatedly to see the effect of edits to your conversion rules: $ watchexec -- "hledger import --dry-run bank.csv | hledger -f- -I print unknown" Once the conversion and dates look good enough to import to your jour- nal, perhaps with some manual fixups to follow, you would do the actual import: $ hledger import bank.csv Overlap detection Reading CSV files is built in to hledger, and not specific to import; so you could also import by doing hledger -f bank.csv print >>$LEDGER_FILE. But import is easier and provides some advantages. The main one is that it avoids re-importing transactions it has seen on previous runs. This means you don't have to worry about overlapping data in successive downloads of your bank CSV; just download and import as often as you like, and only the new transactions will be imported each time. We don't call this "deduplication", as it's generally not possible to reliably detect duplicates in bank CSV. Instead, import remembers the latest date processed previously in each CSV file (saving it in a hid- den file), and skips any records prior to that date. This works well for most real-world CSV, where: 1. the data file name is stable (does not change) across imports 2. the item dates are stable across imports 3. the order of same-date items is stable across imports 4. the newest items have the newest dates (Occasional violations of 2-4 are often harmless; you can reduce the chance of disruption by downloading and importing more often.) Overlap detection is automatic, and shouldn't require much attention from you, except perhaps at first import (see below). But here's how it works: o For each FILE being imported from: 1. hledger reads a file named .latest.FILE file in the same direc- tory, if any. This file contains the latest record date previ- ously imported from FILE, in YYYY-MM-DD format. If multiple records with that date were imported, the date is repeated on N lines. 2. hledger reads records from FILE. If a latest date was found in step 1, any records before that date, and the first N records on that date, are skipped. o After a successful import from all FILEs, without error and without --dry-run, hledger updates each FILE's .latest.FILE for next time. If this goes wrong, it's relatively easy to repair: o You'll notice it before import when you preview with import --dry-run. o Or after import when you try to reconcile your hledger account bal- ances with your bank. o hledger print -f FILE.csv will show all recently downloaded transac- tions. Compare these with your journal. Copy/paste if needed. o Update your conversion rules and print again, if needed. o You can manually update or remove the .latest file, or use import --catchup FILE. o Download and import more often, eg twice a week, at least while you are learning. It's easier to review and troubleshoot when there are fewer transactions. First import The first time you import from a file, when no corresponding .latest file has been created yet, all of the records will be imported. But perhaps you have been entering the data manually, so you know that all of these transactions are already recorded in the journal. In this case you can run hledger import --catchup once. This will create a .latest file containing the latest CSV record date, so that none of those records will be re-imported. Or, if you know that some but not all of the transactions are in the journal, you can create the .latest file yourself. Eg, let's say you previously recorded foobank transactions up to 2024-10-31 in the jour- nal. Then in the directory where you'll be saving foobank.csv, you would create a .latest.foobank.csv file containing 2024-10-31 Or if you had three foobank transactions recorded with that date, you would repeat the date that many times: 2024-10-31 2024-10-31 2024-10-31 Then hledger import foobank.csv [--dry-run] will import only the newer records. Importing balance assignments Journal entries added by import will have all posting amounts made ex- plicit (like print -x). This means that any balance assignments in the imported entries would need to be evaluated. But this generally isn't possible, as the main file's account balances are not visible during import. So try to avoid generating balance assignments with your CSV rules, or importing from a journal that contains balance assignments. (Balance assignments are best avoided anyway.) But if you must use them, eg because your CSV includes only balances: you can import with print, which leaves implicit amounts implicit. (print can also do overlap detection like import, with the --new flag): $ hledger print --new -f bank.csv >> $LEDGER_FILE (If you think import should preserve implicit balances, please test that and send a pull request.) Import and commodity styles Amounts in entries added by import will be formatted according to the journal's canonical commodity styles, as declared by commodity direc- tives or inferred from the journal's amounts. Related: CSV > Amount decimal places. Import archiving When importing from a CSV rules file (hledger import bank.rules), you can use the archive rule to enable automatic archiving of the data file. After a successful import, the data file (specified by source) will be moved to an archive folder (data/, next to the rules file, auto-created), and renamed similar to the rules file, with a date. This can be useful for troubleshooting, detecting variations in your banks' CSV data, regenerating entries with improved rules, etc. The archive rule also causes import to handle source glob patterns dif- ferently: when there are multiple matched files, it will pick the old- est, not the newest. Import special cases Deduplication Here are two kinds of "deduplication" which import does not handle (and should not, because these can happen legitimately in financial data): o Two or more of the new CSV records are identical, and generate iden- tical new journal entries. o A new CSV record generates a journal entry identical to one(s) al- ready in the journal. Varying file name If you have a download whose file name varies, you could rename it to a fixed name after each download. Or you could use a CSV source rule with a suitable glob pattern, and import from the .rules file. Multiple versions Say you download bank.csv, import it, but forget to delete it from your downloads folder. The next time you download it, your web browser will save it as (eg) bank (2).csv. The source rule's glob patterns are for just this situation: instead of specifying source bank.csv, specify source bank*.csv. Then hledger -f bank.rules CMD or hledger import bank.rules will automatically pick the newest matched file (bank (2).csv). Alternately, what if you download, but forget to import or delete, then download again ? Now each of bank.csv and bank (2).csv might contain data that's not in the other, and not in your journal. In this case, it's best to import each of them in turn, oldest first (otherwise, overlap detection could cause new records to be skipped). Enabling im- port archiving ensures this. Then hledger import bank.rules; hledger import bank.rules will import and archive first bank.csv, then bank (2).csv. Basic report commands accounts List the account names used or declared in the journal. Flags: -u --used list accounts used -d --declared list accounts declared --undeclared list accounts used but not declared --unused list accounts declared but not used --find list the first account matched by the first argument (a case-insensitive infix regexp) --types also show account types when known --positions also show where accounts were declared --directives show as account directives, for use in journals -l --flat list/tree mode: show accounts as a flat list (default) -t --tree list/tree mode: show accounts as a tree --drop=N flat mode: omit N leading account name parts This command lists account names - all of them by default. or just the ones which have been used in transactions, or declared with account di- rectives, or used but not declared, or declared but not used, or just the first account name matched by a pattern. You can add query arguments to select a subset of transactions or ac- counts. It shows a flat list by default. With --tree, it uses indentation to show the account hierarchy. In flat mode you can add --drop N to omit the first few account name components. Account names can be depth-clipped with depth:N or --depth N or -N. With --types, it also shows each account's type, if it's known. (See Declaring accounts > Account types.) With --positions, it also shows the file and line number of each ac- count's declaration, if any, and the account's overall declaration or- der; these may be useful when troubleshooting account display order. With --directives, it shows valid account directives which could be pasted into a journal file. This is useful together with --undeclared when updating your account declarations to satisfy hledger check ac- counts. The --find flag can be used to look up a single account name, in the same way that the aregister command does. It returns the alphanumeri- cally-first matched account name, or if none can be found, it fails with a non-zero exit code. Examples: $ hledger accounts assets:bank:checking assets:bank:saving assets:cash expenses:food expenses:supplies income:gifts income:salary liabilities:debts $ hledger accounts --undeclared --directives >> $LEDGER_FILE $ hledger check accounts codes List the codes seen in transactions, in the order parsed. Flags: no command-specific flags This command prints the value of each transaction's code field, in the order transactions were parsed. The transaction code is an optional value written in parentheses between the date and description, often used to store a cheque number, order number or similar. Transactions aren't required to have a code, and missing or empty codes will not be shown by default. With the -E/--empty flag, they will be printed as blank lines. You can add a query to select a subset of transactions. Examples: 2022/1/1 (123) Supermarket Food $5.00 Checking 2022/1/2 (124) Post Office Postage $8.32 Checking 2022/1/3 Supermarket Food $11.23 Checking 2022/1/4 (126) Post Office Postage $3.21 Checking $ hledger codes 123 124 126 $ hledger codes -E 123 124 126 commodities List the commodity symbols used or declared in the journal. Flags: --used list commodities used --declared list commodities declared --undeclared list commodities used but not declared --unused list commodities declared but not used This command lists commodity symbols/names - all of them by default, or just the ones which have been used in transactions or P directives, or declared with commodity directives, or used but not declared, or de- clared but not used. You can add cur: query arguments to further limit the commodities. descriptions List the unique descriptions used in transactions. Flags: no command-specific flags This command lists the unique descriptions that appear in transactions, in alphabetic order. You can add a query to select a subset of trans- actions. Example: $ hledger descriptions Store Name Gas Station | Petrol Person A files List all files included in the journal. With a REGEX argument, only file names matching the regular expression (case sensitive) are shown. Flags: no command-specific flags notes List the unique notes that appear in transactions. Flags: no command-specific flags This command lists the unique notes that appear in transactions, in al- phabetic order. You can add a query to select a subset of transac- tions. The note is the part of the transaction description after a | character (or if there is no |, the whole description). Example: $ hledger notes Petrol Snacks payees List the payee/payer names used or declared in the journal. Flags: --used list payees used --declared list payees declared --undeclared list payees used but not declared --unused list payees declared but not used This command lists unique payee/payer names - all of them by default, or just the ones which have been used in transaction descriptions, or declared with payee directives, or used but not declared, or declared but not used. The payee/payer name is the part of the transaction description before a | character (or if there is no |, the whole description). You can add query arguments to select a subset of transactions or pay- ees. Example: $ hledger payees Store Name Gas Station Person A prices Print the market prices declared with P directives. With --infer-mar- ket-prices, also show any additional prices inferred from costs. With --show-reverse, also show additional prices inferred by reversing known prices. Flags: --show-reverse also show the prices inferred by reversing known prices Price amounts are always displayed with their full precision, except for reverse prices which are limited to 8 decimal digits. Prices can be filtered by a date:, cur: or amt: query. Generally if you run this command with --infer-market-prices --show-re- verse, it will show the same prices used internally to calculate value reports. But if in doubt, you can inspect those directly by running the value report with --debug=2. stats Show journal and performance statistics. Flags: -v --verbose show more detailed output -o --output-file=FILE write output to FILE. The stats command shows summary information for the whole journal, or a matched part of it. With a reporting interval, it shows a report for each report period. The default output is fairly impersonal, though it reveals the main file name. With -v/--verbose, more details are shown, like file paths, included files, and commodity names. It also shows some run time statistics: o elapsed time o throughput: the number of transactions processed per second o live: the peak memory in use by the program to do its work o alloc: the peak memory allocation from the OS as seen by GHC. Mea- suring this externally, eg with GNU time, is more accurate; usually that will be a larger number; sometimes (with swapping?) smaller. The stats command's run time is similar to that of a balance report. Example: $ hledger stats -f examples/1ktxns-1kaccts.journal Main file : .../1ktxns-1kaccts.journal Included files : 0 Txns span : 2000-01-01 to 2002-09-27 (1000 days) Last txn : 2002-09-26 (7827 days ago) Txns : 1000 (1.0 per day) Txns last 30 days : 0 (0.0 per day) Txns last 7 days : 0 (0.0 per day) Payees/descriptions : 1000 Accounts : 1000 (depth 10) Commodities : 26 Market prices : 1000 Runtime stats : 0.12 s elapsed, 8266 txns/s, 4 MB live, 16 MB alloc This command supports the -o/--output-file option (but not -O/--out- put-format). tags List the tag names used or declared in the journal, or their values. Flags: --used list tags used --declared list tags declared --undeclared list tags used but not declared --unused list tags declared but not used --values list tag values instead of tag names --parsed show them in the order they were parsed (mostly), including duplicates This command lists tag names - all of them by default, or just the ones which have been used on transactions/postings/accounts, or declared with tag directives, or used but not declared, or declared but not used. You can add one TAGREGEX argument, to show only tags whose name is matched by this case-insensitive, infix-matching regular expression. After that, you can add query arguments to filter the transactions, postings, or accounts providing tags. With --values, the tags' unique non-empty values are listed instead. With -E/--empty, blank/empty values are also shown. With --parsed, tags or values are shown in the order they were parsed, with duplicates included. (Except, tags from account declarations are always shown first.) Remember that accounts also acquire tags from their parents; postings also acquire tags from their account and transaction; and transactions also acquire tags from their postings. Standard report commands print Show full journal entries, representing transactions. Flags: -x --explicit show all amounts explicitly --show-costs show transaction prices even with conversion postings --round=TYPE how much rounding or padding should be done when displaying amounts ? none - show original decimal digits, as in journal (default) soft - just add or remove decimal zeros to match precision hard - round posting amounts to precision (can unbalance transactions) all - also round cost amounts to precision (can unbalance transactions) --invert display all amounts with reversed sign --new show only newer-dated transactions added in each file since last run -m --match=DESC fuzzy search for one recent transaction with description closest to DESC --base-url=URLPREFIX in html output, generate links to hledger-web, with this prefix. (Usually the base url shown by hledger-web; can also be relative.) --location add tags showing file paths and line numbers -O --output-format=FMT select the output format. Supported formats: txt, beancount, csv, tsv, html, fods, json, sql. -o --output-file=FILE write output to FILE. A file extension matching one of the above formats selects that format. The print command displays full journal entries (transactions) from the journal file, sorted by date (or with --date2, by secondary date). Directives and inter-transaction comments are not shown, currently. This means the print command is somewhat lossy, and if you are using it to reformat/regenerate your journal you should take care to also copy over the directives and inter-transaction comments. Eg: $ hledger print -f examples/sample.journal date:200806 2008/06/01 gift assets:bank:checking $1 income:gifts $-1 2008/06/02 save assets:bank:saving $1 assets:bank:checking $-1 2008/06/03 * eat & shop expenses:food $1 expenses:supplies $1 assets:cash $-2 print explicitness Normally, whether posting amounts are implicit or explicit is pre- served. For example, when an amount is omitted in the journal, it will not appear in the output. Similarly, if a conversion cost is implied but not written, it will not appear in the output. You can use the -x/--explicit flag to force explicit display of all amounts and costs. This can be useful for troubleshooting or for mak- ing your journal more readable and robust against data entry errors. -x is also implied by using any of -B,-V,-X,--value. The -x/--explicit flag will cause any postings with a multi-commodity amount (which can arise when a multi-commodity transaction has an im- plicit amount) to be split into multiple single-commodity postings, keeping the output parseable. print amount style Amounts are shown right-aligned within each transaction (but not aligned across all transactions; you can do that with ledger-mode in Emacs). Amounts will be (mostly) normalised to their commodity display style: their symbol placement, decimal mark, and digit group marks will be made consistent. By default, decimal digits are shown as they are written in the journal. With the --round (Added in 1.32) option, print will try increasingly hard to display decimal digits according to the commodity display styles: o --round=none show amounts with original precisions (default) o --round=soft add/remove decimal zeros in amounts (except costs) o --round=hard round amounts (except costs), possibly hiding signifi- cant digits o --round=all round all amounts and costs soft is good for non-lossy cleanup, formatting amounts more consis- tently where it's safe to do so. hard and all can cause print to show invalid unbalanced journal en- tries; they may be useful eg for stronger cleanup, with manual fixups when needed. print parseability print's output is usually a valid hledger journal, and you can process it again with a second hledger command. This can be useful for certain kinds of search (though the same can be achieved with expr: queries now): # Show running total of food expenses paid from cash. # -f- reads from stdin. -I/--ignore-assertions is sometimes needed. $ hledger print assets:cash | hledger -f- -I reg expenses:food There are some situations where print's output can become unparseable: o Value reporting affects posting amounts but not balance assertion or balance assignment amounts, potentially causing those to fail. o Auto postings can generate postings with too many missing amounts. o Account aliases can generate bad account names. print, other features With -B/--cost, amounts with costs are shown converted to cost. With --invert, posting amounts are shown with their sign flipped. It could be useful if you have accidentally recorded some transactions with the wrong signs. With --new, print shows only transactions it has not seen on a previous run. This uses the same deduplication system as the import command. (See import's docs for details.) With -m DESC/--match=DESC, print shows one recent transaction whose de- scription is most similar to DESC. DESC should contain at least two characters. If there is no similar-enough match, no transaction will be shown and the program exit code will be non-zero. With --location, print adds the source file and line number to every transaction, as a tag. print output format This command also supports the output destination and output format op- tions The output formats supported are txt, beancount (Added in 1.32), csv, tsv (Added in 1.32), json and sql. The beancount format tries to produce Beancount-compatible output, as follows: o Transaction and postings with unmarked status are converted to cleared (*) status. o Transactions' payee and note are backslash-escaped and dou- ble-quote-escaped and wrapped in double quotes. o Transaction tags are copied to Beancount #tag format. o Commodity symbols are converted to upper case, and a small number of currency symbols like $ are converted to the corresponding currency names. o Account name parts are capitalised and unsupported characters are re- placed with -. If an account name part does not begin with a letter, or if the first part is not Assets, Liabilities, Equity, Income, or Expenses, an error is raised. (Use --alias options to bring your ac- counts into compliance.) o An open directive is generated for each account used, on the earliest transaction date. Some limitations: o Balance assertions are removed. o Balance assignments become missing amounts. o Virtual and balanced virtual postings become regular postings. o Directives are not converted. Here's an example of print's CSV output: $ hledger print -Ocsv "txnidx","date","date2","status","code","description","comment","account","amount","commodity","credit","debit","posting-status","posting-comment" "1","2008/01/01","","","","income","","assets:bank:checking","1","$","","1","","" "1","2008/01/01","","","","income","","income:salary","-1","$","1","","","" "2","2008/06/01","","","","gift","","assets:bank:checking","1","$","","1","","" "2","2008/06/01","","","","gift","","income:gifts","-1","$","1","","","" "3","2008/06/02","","","","save","","assets:bank:saving","1","$","","1","","" "3","2008/06/02","","","","save","","assets:bank:checking","-1","$","1","","","" "4","2008/06/03","","*","","eat & shop","","expenses:food","1","$","","1","","" "4","2008/06/03","","*","","eat & shop","","expenses:supplies","1","$","","1","","" "4","2008/06/03","","*","","eat & shop","","assets:cash","-2","$","2","","","" "5","2008/12/31","","*","","pay off","","liabilities:debts","1","$","","1","","" "5","2008/12/31","","*","","pay off","","assets:bank:checking","-1","$","1","","","" o There is one CSV record per posting, with the parent transaction's fields repeated. o The "txnidx" (transaction index) field shows which postings belong to the same transaction. (This number might change if transactions are reordered within the file, files are parsed/included in a different order, etc.) o The amount is separated into "commodity" (the symbol) and "amount" (numeric quantity) fields. o The numeric amount is repeated in either the "credit" or "debit" col- umn, for convenience. (Those names are not accurate in the account- ing sense; it just puts negative amounts under credit and zero or greater amounts under debit.) aregister (areg) Show the transactions and running balances in one account, with each transaction on one line. Flags: --txn-dates filter strictly by transaction date, not posting date. Warning: this can show a wrong running balance. --no-elide don't show only 2 commodities per amount --cumulative accumulation mode: show running total from report start date -H --historical accumulation mode: show historical running total/balance (includes postings before report start date) (default) --invert display all amounts with reversed sign --heading=YN show heading row above table: yes (default) or no -w --width=N set output width (default: terminal width). -wN,M sets description width as well. --align-all guarantee alignment across all lines (slower) -O --output-format=FMT select the output format. Supported formats: txt, html, csv, tsv, json. -o --output-file=FILE write output to FILE. A file extension matching one of the above formats selects that format. aregister shows the overall transactions affecting a particular account (and any subaccounts). Each report line represents one transaction in this account. Transactions before the report start date are included in the running balance (--historical mode is the default). You can suppress this behaviour using the --cumulative option. This is a more "real world", bank-like view than the register command (which shows individual postings, possibly from multiple accounts, not necessarily in historical mode). As a quick rule of thumb: - use areg- ister for reviewing and reconciling real-world asset/liability accounts - use register for reviewing detailed revenues/expenses. aregister requires one argument: the account to report on. You can write either the full account name, or a case-insensitive regular ex- pression which will select the alphabetically first matched account. When there are multiple matches, the alphabetically-first choice can be surprising; eg if you have assets:per:checking 1 and assets:biz:check- ing 2 accounts, hledger areg checking would select assets:biz:checking 2. It's just a convenience to save typing, so if in doubt, write the full account name, or a distinctive substring that matches uniquely. Transactions involving subaccounts of this account will also be shown. aregister ignores depth limits, so its final total will always match a historical balance report with similar arguments. Any additional arguments form a query which will filter the transac- tions shown. Note some queries will disturb the running balance, caus- ing it to be different from the account's real-world running balance. An example: this shows the transactions and historical running balance during july, in the first account whose name contains "checking": $ hledger areg checking date:jul Each aregister line item shows: o the transaction's date (or the relevant posting's date if different, see below) o the names of all the other account(s) involved in this transaction (probably abbreviated) o the total change to this account's balance from this transaction o the account's historical running balance after this transaction. Transactions making a net change of zero are not shown by default; add the -E/--empty flag to show them. For performance reasons, column widths are chosen based on the first 1000 lines; this means unusually wide values in later lines can cause visual discontinuities as column widths are adjusted. If you want to ensure perfect alignment, at the cost of more time and memory, use the --align-all flag. By default, aregister shows a heading above the data. However, when reporting in a language different from English, it is easier to omit this heading and prepend your own one. For this purpose, use the --heading=no option. This command also supports the output destination and output format op- tions. The output formats supported are txt, csv, tsv (Added in 1.32), html, fods (Added in 1.41) and json. aregister and posting dates aregister always shows one line (and date and amount) per transaction. But sometimes transactions have postings with different dates. Also, not all of a transaction's postings may be within the report period. To resolve this, aregister shows the earliest of the transaction's date and posting dates that is in-period, and the sum of the in-period post- ings. In other words it will show a combined line item with just the earliest date, and the running balance will (temporarily, until the transaction's last posting) be inaccurate. Use register -H if you need to see the individual postings. There is also a --txn-dates flag, which filters strictly by transaction date, ignoring posting dates. This too can cause an inaccurate running balance. register (reg) Show postings and their running total. Flags: --cumulative accumulation mode: show running total from report start date (default) -H --historical accumulation mode: show historical running total/balance (includes postings before report start date) -A --average show running average of posting amounts instead of total (implies --empty) -m --match=DESC fuzzy search for one recent posting with description closest to DESC -r --related show postings' siblings instead --invert display all amounts with reversed sign --sort=FIELDS sort by: date, desc, account, amount, absamount, or a comma-separated combination of these. For a descending sort, prefix with -. (Default: date) -w --width=N set output width (default: terminal width). -wN,M sets description width as well. --align-all guarantee alignment across all lines (slower) --base-url=URLPREFIX in html output, generate links to hledger-web, with this prefix. (Usually the base url shown by hledger-web; can also be relative.) -O --output-format=FMT select the output format. Supported formats: txt, csv, tsv, html, fods, json. -o --output-file=FILE write output to FILE. A file extension matching one of the above formats selects that format. The register command displays matched postings, across all accounts, in date order, with their running total or running historical balance. (See also the aregister command, which shows matched transactions in a specific account.) register normally shows line per posting, but note that multi-commodity amounts will occupy multiple lines (one line per commodity). It is typically used with a query selecting a particular account, to see that account's activity: $ hledger register checking 2008/01/01 income assets:bank:checking $1 $1 2008/06/01 gift assets:bank:checking $1 $2 2008/06/02 save assets:bank:checking $-1 $1 2008/12/31 pay off assets:bank:checking $-1 0 With --date2, it shows and sorts by secondary date instead. For performance reasons, column widths are chosen based on the first 1000 lines; this means unusually wide values in later lines can cause visual discontinuities as column widths are adjusted. If you want to ensure perfect alignment, at the cost of more time and memory, use the --align-all flag. The --historical/-H flag adds the balance from any undisplayed prior postings to the running total. This is useful when you want to see only recent activity, with a historically accurate running balance: $ hledger register checking -b 2008/6 --historical 2008/06/01 gift assets:bank:checking $1 $2 2008/06/02 save assets:bank:checking $-1 $1 2008/12/31 pay off assets:bank:checking $-1 0 The --depth option limits the amount of sub-account detail displayed. The --average/-A flag shows the running average posting amount instead of the running total (so, the final number displayed is the average for the whole report period). This flag implies --empty (see below). It is affected by --historical. It works best when showing just one ac- count and one commodity. The --related/-r flag shows the other postings in the transactions of the postings which would normally be shown. The --invert flag negates all amounts. For example, it can be used on an income account where amounts are normally displayed as negative num- bers. It's also useful to show postings on the checking account to- gether with the related account: The --sort=FIELDS flag sorts by the fields given, which can be any of account, amount, absamount, date, or desc/description, optionally sepa- rated by commas. For example, --sort account,amount will group all transactions in each account, sorted by transaction amount. Each field can be negated by a preceding -, so --sort -amount will show transac- tions ordered from smallest amount to largest amount. $ hledger register --related --invert assets:checking With a reporting interval, register shows summary postings, one per in- terval, aggregating the postings to each account: $ hledger register --monthly income 2008/01 income:salary $-1 $-1 2008/06 income:gifts $-1 $-2 Periods with no activity, and summary postings with a zero amount, are not shown by default; use the --empty/-E flag to see them: $ hledger register --monthly income -E 2008/01 income:salary $-1 $-1 2008/02 0 $-1 2008/03 0 $-1 2008/04 0 $-1 2008/05 0 $-1 2008/06 income:gifts $-1 $-2 2008/07 0 $-2 2008/08 0 $-2 2008/09 0 $-2 2008/10 0 $-2 2008/11 0 $-2 2008/12 0 $-2 Often, you'll want to see just one line per interval. The --depth op- tion helps with this, causing subaccounts to be aggregated: $ hledger register --monthly assets --depth 1h 2008/01 assets $1 $1 2008/06 assets $-1 0 2008/12 assets $-1 $-1 Note when using report intervals, if you specify start/end dates these will be adjusted outward if necessary to contain a whole number of in- tervals. This ensures that the first and last intervals are full length and comparable to the others in the report. With -m DESC/--match=DESC, register does a fuzzy search for one recent posting whose description is most similar to DESC. DESC should contain at least two characters. If there is no similar-enough match, no post- ing will be shown and the program exit code will be non-zero. Custom register output register normally uses the full terminal width (or 80 columns if it can't detect that). You can override this with the --width/-w option. The description and account columns normally share the space equally (about half of (width - 40) each). You can adjust this by adding a de- scription width as part of --width's argument, comma-separated: --width W,D . Here's a diagram (won't display correctly in --help): <--------------------------------- width (W) ----------------------------------> date (10) description (D) account (W-41-D) amount (12) balance (12) DDDDDDDDDD dddddddddddddddddddd aaaaaaaaaaaaaaaaaaa AAAAAAAAAAAA AAAAAAAAAAAA and some examples: $ hledger reg # use terminal width (or 80 on windows) $ hledger reg -w 100 # use width 100 $ hledger reg -w 100,40 # set overall width 100, description width 40 This command also supports the output destination and output format op- tions The output formats supported are txt, csv, tsv (Added in 1.32), and json. balancesheet (bs) Show the end balances in asset and liability accounts. Amounts are shown with normal positive sign, as in conventional financial state- ments. Flags: --sum calculation mode: show sum of posting amounts (default) --valuechange calculation mode: show total change of value of period-end historical balances (caused by deposits, withdrawals, market price fluctuations) --gain calculation mode: show unrealised capital gain/loss (historical balance value minus cost basis) --count calculation mode: show the count of postings --change accumulation mode: accumulate amounts from column start to column end (in multicolumn reports) --cumulative accumulation mode: accumulate amounts from report start (specified by e.g. -b/--begin) to column end -H --historical accumulation mode: accumulate amounts from journal start to column end (includes postings before report start date) (default) -l --flat list/tree mode: show accounts as a flat list (default). Amounts exclude subaccount amounts, except where the account is depth-clipped. -t --tree list/tree mode: show accounts as a tree. Amounts include subaccount amounts. --drop=N in list mode, omit N leading account name parts --declared include non-parent declared accounts (best used with -E) -A --average show a row average column (in multicolumn reports) -T --row-total show a row total column (in multicolumn reports) --summary-only display only row summaries (e.g. row total, average) (in multicolumn reports) -N --no-total omit the final total row --no-elide in tree mode, don't squash boring parent accounts --format=FORMATSTR use this custom line format (in simple reports) -S --sort-amount sort by amount instead of account code/name -% --percent express values in percentage of each column's total --layout=ARG how to show multi-commodity amounts: 'wide[,WIDTH]': all commodities on one line 'tall' : each commodity on a new line 'bare' : bare numbers, symbols in a column --base-url=URLPREFIX in html output, generate hyperlinks to hledger-web, with this prefix. (Usually the base url shown by hledger-web; can also be relative.) -O --output-format=FMT select the output format. Supported formats: txt, html, csv, tsv, json. -o --output-file=FILE write output to FILE. A file extension matching one of the above formats selects that format. This command displays a balance sheet, showing historical ending bal- ances of asset and liability accounts. (To see equity as well, use the balancesheetequity command.) Accounts declared with the Asset, Cash or Liability type are shown (see account types). Or if no such accounts are declared, it shows top-level accounts named asset or liability (case insensitive, plurals allowed) and their subaccounts. Example: $ hledger balancesheet Balance Sheet 2008-12-31 || 2008-12-31 ====================++============ Assets || --------------------++------------ assets:bank:saving || $1 assets:cash || $-2 --------------------++------------ || $-1 ====================++============ Liabilities || --------------------++------------ liabilities:debts || $-1 --------------------++------------ || $-1 ====================++============ Net: || 0 This command is a higher-level variant of the balance command, and sup- ports many of that command's features, such as multi-period reports. It is similar to hledger balance -H assets liabilities, but with smarter account detection, and liabilities displayed with their sign flipped. This command also supports the output destination and output format op- tions The output formats supported are txt, csv, tsv (Added in 1.32), html, and json. balancesheetequity (bse) This command displays a balance sheet, showing historical ending bal- ances of asset, liability and equity accounts. Amounts are shown with normal positive sign, as in conventional financial statements. Flags: --sum calculation mode: show sum of posting amounts (default) --valuechange calculation mode: show total change of value of period-end historical balances (caused by deposits, withdrawals, market price fluctuations) --gain calculation mode: show unrealised capital gain/loss (historical balance value minus cost basis) --count calculation mode: show the count of postings --change accumulation mode: accumulate amounts from column start to column end (in multicolumn reports) --cumulative accumulation mode: accumulate amounts from report start (specified by e.g. -b/--begin) to column end -H --historical accumulation mode: accumulate amounts from journal start to column end (includes postings before report start date) (default) -l --flat list/tree mode: show accounts as a flat list (default). Amounts exclude subaccount amounts, except where the account is depth-clipped. -t --tree list/tree mode: show accounts as a tree. Amounts include subaccount amounts. --drop=N in list mode, omit N leading account name parts --declared include non-parent declared accounts (best used with -E) -A --average show a row average column (in multicolumn reports) -T --row-total show a row total column (in multicolumn reports) --summary-only display only row summaries (e.g. row total, average) (in multicolumn reports) -N --no-total omit the final total row --no-elide in tree mode, don't squash boring parent accounts --format=FORMATSTR use this custom line format (in simple reports) -S --sort-amount sort by amount instead of account code/name -% --percent express values in percentage of each column's total --layout=ARG how to show multi-commodity amounts: 'wide[,WIDTH]': all commodities on one line 'tall' : each commodity on a new line 'bare' : bare numbers, symbols in a column --base-url=URLPREFIX in html output, generate hyperlinks to hledger-web, with this prefix. (Usually the base url shown by hledger-web; can also be relative.) -O --output-format=FMT select the output format. Supported formats: txt, html, csv, tsv, json. -o --output-file=FILE write output to FILE. A file extension matching one of the above formats selects that format. This report shows accounts declared with the Asset, Cash, Liability or Equity type (see account types). Or if no such accounts are declared, it shows top-level accounts named asset, liability or equity (case in- sensitive, plurals allowed) and their subaccounts. Example: $ hledger balancesheetequity Balance Sheet With Equity 2008-12-31 || 2008-12-31 ====================++============ Assets || --------------------++------------ assets:bank:saving || $1 assets:cash || $-2 --------------------++------------ || $-1 ====================++============ Liabilities || --------------------++------------ liabilities:debts || $-1 --------------------++------------ || $-1 ====================++============ Equity || --------------------++------------ --------------------++------------ || 0 ====================++============ Net: || 0 This command is a higher-level variant of the balance command, and sup- ports many of that command's features, such as multi-period reports. It is similar to hledger balance -H assets liabilities equity, but with smarter account detection, and liabilities/equity displayed with their sign flipped. This report is the easiest way to see if the accounting equation (A+L+E = 0) is satisfied (after you have done a close --retain to merge rev- enues and expenses with equity, and perhaps added --infer-equity to balance your commodity conversions). This command also supports the output destination and output format op- tions The output formats supported are txt, csv, tsv, html, and json. cashflow (cf) This command displays a (simple) cashflow statement, showing the in- flows and outflows affecting "cash" (ie, liquid, easily convertible) assets. Amounts are shown with normal positive sign, as in conven- tional financial statements. Flags: --sum calculation mode: show sum of posting amounts (default) --valuechange calculation mode: show total change of value of period-end historical balances (caused by deposits, withdrawals, market price fluctuations) --gain calculation mode: show unrealised capital gain/loss (historical balance value minus cost basis) --count calculation mode: show the count of postings --change accumulation mode: accumulate amounts from column start to column end (in multicolumn reports) (default) --cumulative accumulation mode: accumulate amounts from report start (specified by e.g. -b/--begin) to column end -H --historical accumulation mode: accumulate amounts from journal start to column end (includes postings before report start date) -l --flat list/tree mode: show accounts as a flat list (default). Amounts exclude subaccount amounts, except where the account is depth-clipped. -t --tree list/tree mode: show accounts as a tree. Amounts include subaccount amounts. --drop=N in list mode, omit N leading account name parts --declared include non-parent declared accounts (best used with -E) -A --average show a row average column (in multicolumn reports) -T --row-total show a row total column (in multicolumn reports) --summary-only display only row summaries (e.g. row total, average) (in multicolumn reports) -N --no-total omit the final total row --no-elide in tree mode, don't squash boring parent accounts --format=FORMATSTR use this custom line format (in simple reports) -S --sort-amount sort by amount instead of account code/name -% --percent express values in percentage of each column's total --layout=ARG how to show multi-commodity amounts: 'wide[,WIDTH]': all commodities on one line 'tall' : each commodity on a new line 'bare' : bare numbers, symbols in a column --base-url=URLPREFIX in html output, generate hyperlinks to hledger-web, with this prefix. (Usually the base url shown by hledger-web; can also be relative.) -O --output-format=FMT select the output format. Supported formats: txt, html, csv, tsv, json. -o --output-file=FILE write output to FILE. A file extension matching one of the above formats selects that format. This report shows accounts declared with the Cash type (see account types). Or if no such accounts are declared, it shows accounts o under a top-level account named asset (case insensitive, plural al- lowed) o whose name contains some variation of cash, bank, checking or saving. More precisely: all accounts matching this case insensitive regular ex- pression: ^assets?(:.+)?:(cash|bank|che(ck|que?)(ing)?|savings?|currentcash)(:|$) and their subaccounts. An example cashflow report: $ hledger cashflow Cashflow Statement 2008 || 2008 ====================++====== Cash flows || --------------------++------ assets:bank:saving || $1 assets:cash || $-2 --------------------++------ || $-1 This command is a higher-level variant of the balance command, and sup- ports many of that command's features, such as multi-period reports. It is similar to hledger balance assets not:fixed not:investment not:receivable, but with smarter account detection. This command also supports the output destination and output format op- tions The output formats supported are txt, csv, tsv (Added in 1.32), html, and json. incomestatement (is) Show revenue inflows and expense outflows during the report period. Amounts are shown with normal positive sign, as in conventional finan- cial statements. Flags: --sum calculation mode: show sum of posting amounts (default) --valuechange calculation mode: show total change of value of period-end historical balances (caused by deposits, withdrawals, market price fluctuations) --gain calculation mode: show unrealised capital gain/loss (historical balance value minus cost basis) --count calculation mode: show the count of postings --change accumulation mode: accumulate amounts from column start to column end (in multicolumn reports) (default) --cumulative accumulation mode: accumulate amounts from report start (specified by e.g. -b/--begin) to column end -H --historical accumulation mode: accumulate amounts from journal start to column end (includes postings before report start date) -l --flat list/tree mode: show accounts as a flat list (default). Amounts exclude subaccount amounts, except where the account is depth-clipped. -t --tree list/tree mode: show accounts as a tree. Amounts include subaccount amounts. --drop=N in list mode, omit N leading account name parts --declared include non-parent declared accounts (best used with -E) -A --average show a row average column (in multicolumn reports) -T --row-total show a row total column (in multicolumn reports) --summary-only display only row summaries (e.g. row total, average) (in multicolumn reports) -N --no-total omit the final total row --no-elide in tree mode, don't squash boring parent accounts --format=FORMATSTR use this custom line format (in simple reports) -S --sort-amount sort by amount instead of account code/name -% --percent express values in percentage of each column's total --layout=ARG how to show multi-commodity amounts: 'wide[,WIDTH]': all commodities on one line 'tall' : each commodity on a new line 'bare' : bare numbers, symbols in a column --base-url=URLPREFIX in html output, generate hyperlinks to hledger-web, with this prefix. (Usually the base url shown by hledger-web; can also be relative.) -O --output-format=FMT select the output format. Supported formats: txt, html, csv, tsv, json. -o --output-file=FILE write output to FILE. A file extension matching one of the above formats selects that format. This command displays an income statement, showing revenues and ex- penses during one or more periods. It shows accounts declared with the Revenue or Expense type (see ac- count types). Or if no such accounts are declared, it shows top-level accounts named revenue or income or expense (case insensitive, plurals allowed) and their subaccounts. Example: $ hledger incomestatement Income Statement 2008 || 2008 ===================++====== Revenues || -------------------++------ income:gifts || $1 income:salary || $1 -------------------++------ || $2 ===================++====== Expenses || -------------------++------ expenses:food || $1 expenses:supplies || $1 -------------------++------ || $2 ===================++====== Net: || 0 This command is a higher-level variant of the balance command, and sup- ports many of that command's features, such as multi-period reports. It is similar to hledger balance '(revenues|income)' expenses, but with smarter account detection, and revenues/income displayed with their sign flipped. This command also supports the output destination and output format op- tions The output formats supported are txt, csv, tsv (Added in 1.32), html, and json. Advanced report commands balance (bal) A flexible, general purpose "summing" report that shows accounts with some kind of numeric data. This can be balance changes per period, end balances, budget performance, unrealised capital gains, etc. Flags: --sum calculation mode: show sum of posting amounts (default) --valuechange calculation mode: show total change of value of period-end historical balances (caused by deposits, withdrawals, market price fluctuations) --gain calculation mode: show unrealised capital gain/loss (historical balance value minus cost basis) --budget[=DESCPAT] calculation mode: show sum of posting amounts together with budget goals defined by periodic transactions. With a DESCPAT argument (must be separated by = not space), use only periodic transactions with matching description (case insensitive substring match). --count calculation mode: show the count of postings --change accumulation mode: accumulate amounts from column start to column end (in multicolumn reports, default) --cumulative accumulation mode: accumulate amounts from report start (specified by e.g. -b/--begin) to column end -H --historical accumulation mode: accumulate amounts from journal start to column end (includes postings before report start date) -l --flat list/tree mode: show accounts as a flat list (default). Amounts exclude subaccount amounts, except where the account is depth-clipped. -t --tree list/tree mode: show accounts as a tree. Amounts include subaccount amounts. --drop=N in list mode, omit N leading account name parts --declared include non-parent declared accounts (best used with -E) -A --average show a row average column (in multicolumn reports) -T --row-total show a row total column (in multicolumn reports) --summary-only display only row summaries (e.g. row total, average) (in multicolumn reports) -N --no-total omit the final total row --no-elide in tree mode, don't squash boring parent accounts --format=FORMATSTR use this custom line format (in simple reports) -S --sort-amount sort by amount instead of account code/name (in flat mode). With multiple columns, sorts by the row total, or by row average if that is displayed. -% --percent express values in percentage of each column's total -r --related show the other accounts transacted with, instead --invert display all amounts with reversed sign --transpose switch rows and columns (use vertical time axis) --layout=ARG how to lay out multi-commodity amounts and the overall table: 'wide[,W]': commodities on same line, up to W wide 'tall' : commodities on separate lines 'bare' : commodity symbols in a separate column 'tidy' : each data field in its own column --base-url=URLPREFIX in html output, generate links to hledger-web, with this prefix. (Usually the base url shown by hledger-web; can also be relative.) -O --output-format=FMT select the output format. Supported formats: txt, html, csv, tsv, json, fods. -o --output-file=FILE write output to FILE. A file extension matching one of the above formats selects that format. balance is one of hledger's oldest and most versatile commands, for listing account balances, balance changes, values, value changes and more, during one time period or many. Generally it shows a table, with rows representing accounts, and columns representing periods. Note there are some variants of the balance command with convenient de- faults, which are simpler to use: balancesheet, balancesheetequity, cashflow and incomestatement. When you need more control, then use balance. balance features Here's a quick overview of the balance command's features, followed by more detailed descriptions and examples. Many of these work with the other balance-like commands as well (bs, cf, is..). balance can show.. o accounts as a list (-l) or a tree (-t) o optionally depth-limited (-[1-9]) o sorted by declaration order and name, or by amount ..and their.. o balance changes (the default) o or actual and planned balance changes (--budget) o or value of balance changes (-V) o or change of balance values (--valuechange) o or unrealised capital gain/loss (--gain) o or balance changes from sibling postings (--related/-r) o or postings count (--count) ..in.. o one time period (the whole journal period by default) o or multiple periods (-D, -W, -M, -Q, -Y, -p INTERVAL) ..either.. o per period (the default) o or accumulated since report start date (--cumulative) o or accumulated since account creation (--historical/-H) ..possibly converted to.. o cost (--value=cost[,COMM]/--cost/-B) o or market value, as of transaction dates (--value=then[,COMM]) o or at period ends (--value=end[,COMM]) o or now (--value=now) o or at some other date (--value=YYYY-MM-DD) ..with.. o totals (-T), averages (-A), percentages (-%), inverted sign (--in- vert) o rows and columns swapped (--transpose) o another field used as account name (--pivot) o custom-formatted line items (single-period reports only) (--format) o commodities displayed on the same line or multiple lines (--layout) This command supports the output destination and output format options, with output formats txt, csv, tsv (Added in 1.32), json, and (multi-pe- riod reports only:) html, fods (Added in 1.40). In txt output in a colour-supporting terminal, negative amounts are shown in red. Simple balance report With no arguments, balance shows a list of all accounts and their change of balance - ie, the sum of posting amounts, both inflows and outflows - during the entire period of the journal. ("Simple" here means just one column of numbers, covering a single period. You can also have multi-period reports, described later.) For real-world accounts, these numbers will normally be their end bal- ance at the end of the journal period; more on this below. Accounts are sorted by declaration order if any, and then alphabeti- cally by account name. For instance (using examples/sample.journal): $ hledger -f examples/sample.journal bal $1 assets:bank:saving $-2 assets:cash $1 expenses:food $1 expenses:supplies $-1 income:gifts $-1 income:salary $1 liabilities:debts -------------------- 0 Accounts with a zero balance (and no non-zero subaccounts, in tree mode - see below) are hidden by default. Use -E/--empty to show them (re- vealing assets:bank:checking here): $ hledger -f examples/sample.journal bal -E 0 assets:bank:checking $1 assets:bank:saving $-2 assets:cash $1 expenses:food $1 expenses:supplies $-1 income:gifts $-1 income:salary $1 liabilities:debts -------------------- 0 The total of the amounts displayed is shown as the last line, unless -N/--no-total is used. Balance report line format For single-period balance reports displayed in the terminal (only), you can use --format FMT to customise the format and content of each line. Eg: $ hledger -f examples/sample.journal balance --format "%20(account) %12(total)" assets $-1 bank:saving $1 cash $-2 expenses $2 food $1 supplies $1 income $-2 gifts $-1 salary $-1 liabilities:debts $1 --------------------------------- 0 The FMT format string specifies the formatting applied to each ac- count/balance pair. It may contain any suitable text, with data fields interpolated like so: %[MIN][.MAX](FIELDNAME) o MIN pads with spaces to at least this width (optional) o MAX truncates at this width (optional) o FIELDNAME must be enclosed in parentheses, and can be one of: o depth_spacer - a number of spaces equal to the account's depth, or if MIN is specified, MIN * depth spaces. o account - the account's name o total - the account's balance/posted total, right justified Also, FMT can begin with an optional prefix to control how multi-com- modity amounts are rendered: o %_ - render on multiple lines, bottom-aligned (the default) o %^ - render on multiple lines, top-aligned o %, - render on one line, comma-separated There are some quirks. Eg in one-line mode, %(depth_spacer) has no ef- fect, instead %(account) has indentation built in. Experimentation may be needed to get pleasing results. Some example formats: o %(total) - the account's total o %-20.20(account) - the account's name, left justified, padded to 20 characters and clipped at 20 characters o %,%-50(account) %25(total) - account name padded to 50 characters, total padded to 20 characters, with multiple commodities rendered on one line o %20(total) %2(depth_spacer)%-(account) - the default format for the single-column balance report Filtered balance report You can show fewer accounts, a different time period, totals from cleared transactions only, etc. by using query arguments or options to limit the postings being matched. Eg: $ hledger -f examples/sample.journal bal --cleared assets date:200806 $-2 assets:cash -------------------- $-2 List or tree mode By default, or with -l/--flat, accounts are shown as a flat list with their full names visible, as in the examples above. With -t/--tree, the account hierarchy is shown, with subaccounts' "leaf" names indented below their parent: $ hledger -f examples/sample.journal balance $-1 assets $1 bank:saving $-2 cash $2 expenses $1 food $1 supplies $-2 income $-1 gifts $-1 salary $1 liabilities:debts -------------------- 0 Notes: o "Boring" accounts are combined with their subaccount for more compact output, unless --no-elide is used. Boring accounts have no balance of their own and just one subaccount (eg assets:bank and liabilities above). o All balances shown are "inclusive", ie including the balances from all subaccounts. Note this means some repetition in the output, which requires explanation when sharing reports with non-plaintextac- counting-users. A tree mode report's final total is the sum of the top-level balances shown, not of all the balances shown. o Each group of sibling accounts (ie, under a common parent) is sorted separately. Depth limiting With a depth:NUM query, or --depth NUM option, or just -NUM (eg: -3) balance reports will show accounts only to the specified depth, hiding the deeper subaccounts. This can be useful for getting an overview without too much detail. Account balances at the depth limit always include the balances from any deeper subaccounts (even in list mode). Eg, limiting to depth 1: $ hledger -f examples/sample.journal balance -1 $-1 assets $2 expenses $-2 income $1 liabilities -------------------- 0 Dropping top-level accounts You can also hide one or more top-level account name parts, using --drop NUM. This can be useful for hiding repetitive top-level account names: $ hledger -f examples/sample.journal bal expenses --drop 1 $1 food $1 supplies -------------------- $2 Showing declared accounts With --declared, accounts which have been declared with an account di- rective will be included in the balance report, even if they have no transactions. (Since they will have a zero balance, you will also need -E/--empty to see them.) More precisely, leaf declared accounts (with no subaccounts) will be included, since those are usually the more useful in reports. The idea of this is to be able to see a useful "complete" balance re- port, even when you don't have transactions in all of your declared ac- counts yet. Sorting by amount With -S/--sort-amount, accounts with the largest (most positive) bal- ances are shown first. Eg: hledger bal expenses -MAS shows your biggest averaged monthly expenses first. When more than one commodity is present, they will be sorted by the alphabetically earliest commod- ity first, and then by subsequent commodities (if an amount is missing a commodity, it is treated as 0). Revenues and liability balances are typically negative, however, so -S shows these in reverse order. To work around this, you can add --in- vert to flip the signs. Or you could use one of the higher-level bal- ance reports (bs, is..), which flip the sign automatically (eg: hledger is -MAS). Percentages With -%/--percent, balance reports show each account's value expressed as a percentage of the (column) total. Note it is not useful to calculate percentages if the amounts in a col- umn have mixed signs. In this case, make a separate report for each sign, eg: $ hledger bal -% amt:`>0` $ hledger bal -% amt:`<0` Similarly, if the amounts in a column have mixed commodities, convert them to one commodity with -B, -V, -X or --value, or make a separate report for each commodity: $ hledger bal -% cur:\\$ $ hledger bal -% cur: Multi-period balance report With a report interval (set by the -D/--daily, -W/--weekly, -M/--monthly, -Q/--quarterly, -Y/--yearly, or -p/--period flag), bal- ance shows a tabular report, with columns representing successive time periods (and a title): $ hledger -f examples/sample.journal bal --quarterly income expenses -E Balance changes in 2008: || 2008q1 2008q2 2008q3 2008q4 ===================++================================= expenses:food || 0 $1 0 0 expenses:supplies || 0 $1 0 0 income:gifts || 0 $-1 0 0 income:salary || $-1 0 0 0 -------------------++--------------------------------- || $-1 $1 0 0 Notes: o The report's start/end dates will be expanded, if necessary, to fully encompass the displayed subperiods (so that the first and last subpe- riods have the same duration as the others). o Leading and trailing periods (columns) containing all zeroes are not shown, unless -E/--empty is used. o Accounts (rows) containing all zeroes are not shown, unless -E/--empty is used. o Amounts with many commodities are shown in abbreviated form, unless --no-elide is used. o Average and/or total columns can be added with the -A/--average and -T/--row-total flags. o The --transpose flag can be used to exchange rows and columns. o The --pivot FIELD option causes a different transaction field to be used as "account name". See PIVOTING. o The --summary-only flag (--summary also works) hides all but the To- tal and Average columns (those should be enabled with --row-total and -A/--average). Multi-period reports with many periods can be too wide for easy viewing in the terminal. Here are some ways to handle that: o Hide the totals row with -N/--no-total o Filter to a single currency with cur: o Convert to a single currency with -V [--infer-market-price] o Use a more compact layout like --layout=bare o Maximize the terminal window o Reduce the terminal's font size o View with a pager like less, eg: hledger bal -D --color=yes | less -RS o Output as CSV and use a CSV viewer like visidata (hledger bal -D -O csv | vd -f csv), Emacs' csv-mode (M-x csv-mode, C-c C-a), or a spreadsheet (hledger bal -D -o a.csv && open a.csv) o Output as HTML and view with a browser: hledger bal -D -o a.html && open a.html Balance change, end balance It's important to be clear on the meaning of the numbers shown in bal- ance reports. Here is some terminology we use: A balance change is the net amount added to, or removed from, an ac- count during some period. An end balance is the amount accumulated in an account as of some date (and some time, but hledger doesn't store that; assume end of day in your timezone). It is the sum of previous balance changes. We call it a historical end balance if it includes all balance changes since the account was created. For a real world account, this means it will match the "historical record", eg the balances reported in your bank statements or bank web UI. (If they are correct!) In general, balance changes are what you want to see when reviewing revenues and expenses, and historical end balances are what you want to see when reviewing or reconciling asset, liability and equity accounts. balance shows balance changes by default. To see accurate historical end balances: 1. Initialise account starting balances with an "opening balances" transaction (a transfer from equity to the account), unless the journal covers the account's full lifetime. 2. Include all of of the account's prior postings in the report, by not specifying a report start date, or by using the -H/--historical flag. (-H causes report start date to be ignored when summing post- ings.) Balance report modes The balance command is quite flexible; here is the full detail on how to control what it reports. If the following seems complicated, don't worry - this is for advanced reporting, and it does take time and ex- perimentation to get familiar with all the report modes. There are three important option groups: hledger balance [CALCULATIONMODE] [ACCUMULATIONMODE] [VALUATIONMODE] ... Calculation mode The basic calculation to perform for each table cell. It is one of: o --sum : sum the posting amounts (default) o --budget : sum the amounts, but also show the budget goal amount (for each account/period) o --valuechange : show the change in period-end historical balance val- ues (caused by deposits, withdrawals, and/or market price fluctua- tions) o --gain : show the unrealised capital gain/loss, (the current valued balance minus each amount's original cost) o --count : show the count of postings Accumulation mode How amounts should accumulate across a report's subperiods/columns. Another way to say it: which time period's postings should contribute to each cell's calculation. It is one of: o --change : calculate with postings from column start to column end, ie "just this column". Typically used to see revenues/expenses. (default for balance, cashflow, incomestatement) o --cumulative : calculate with postings from report start to column end, ie "previous columns plus this column". Typically used to show changes accumulated since the report's start date. Not often used. o --historical/-H : calculate with postings from journal start to col- umn end, ie "all postings from before report start date until this column's end". Typically used to see historical end balances of as- sets/liabilities/equity. (default for balancesheet, balancesheete- quity) Valuation mode Which kind of value or cost conversion should be applied, if any, be- fore displaying the report. See Cost reporting and Value reporting for more about conversions. A valuation (or cost) mode can be selected with the --value option: o no conversion : don't convert to cost or value (default) o --value=cost[,COMM] : convert amounts to cost (then optionally to some other commodity) o --value=then[,COMM] : convert amounts to market value on transaction dates o --value=end[,COMM] : convert amounts to market value on period end date(s) (default with --valuechange, --gain) o --value=now[,COMM] : convert amounts to market value on today's date o --value=YYYY-MM-DD[,COMM] : convert amounts to market value on an- other date or with the legacy -B/-V/-X options, which are equivalent and easier to type: o -B/--cost : like --value=cost o -V/--market : like --value=end o -X COMM/--exchange COMM : like --value=end,COMM Note that --value can also convert to cost, as a convenience; but actu- ally --cost and --value are independent options, and could be used to- gether. Combining balance report modes Most combinations of these modes should produce reasonable reports, but if you find any that seem wrong or misleading, let us know. The fol- lowing restrictions are applied: o --valuechange implies --value=end o --valuechange makes --change the default when used with the bal- ancesheet/balancesheetequity commands o --cumulative or --historical disables --row-total/-T For reference, here is what the combinations of accumulation and valua- tion show: Valua- no valuation --value= then --value= end --value= tion:> YYYY-MM-DD Accumu- /now lation:v ----------------------------------------------------------------------------------- --change change in period sum of post- period-end DATE-value of ing-date market value of change change in pe- values in period in period riod --cumu- change from re- sum of post- period-end DATE-value of lative port start to ing-date market value of change change from period end values from re- from report report start port start to pe- start to period to period end riod end end --his- change from sum of post- period-end DATE-value of torical journal start to ing-date market value of change change from /-H period end (his- values from jour- from journal journal start torical end bal- nal start to pe- start to period to period end ance) riod end end Budget report The --budget report is like a regular balance report, but with two main differences: o Budget goals and performance percentages are also shown, in brackets o Accounts which don't have budget goals are hidden by default. This is useful for comparing planned and actual income, expenses, time usage, etc. Periodic transaction rules are used to define budget goals. For exam- ple, here's a periodic rule defining monthly goals for bus travel and food expenses: ;; Budget ~ monthly (expenses:bus) $30 (expenses:food) $400 After recording some actual expenses, ;; Two months worth of expenses 2017-11-01 income $-1950 expenses:bus $35 expenses:food:groceries $310 expenses:food:dining $42 expenses:movies $38 assets:bank:checking 2017-12-01 income $-2100 expenses:bus $53 expenses:food:groceries $380 expenses:food:dining $32 expenses:gifts $100 assets:bank:checking we can see a budget report like this: $ hledger bal -M --budget Budget performance in 2017-11-01..2017-12-31: || Nov Dec ===============++============================================ || $-425 $-565 expenses || $425 [ 99% of $430] $565 [131% of $430] expenses:bus || $35 [117% of $30] $53 [177% of $30] expenses:food || $352 [ 88% of $400] $412 [103% of $400] ---------------++-------------------------------------------- || 0 [ 0% of $430] 0 [ 0% of $430] This is "goal-based budgeting"; you define goals for accounts and peri- ods, often recurring, and hledger shows performance relative to the goals. This contrasts with "envelope budgeting", which is more de- tailed and strict - useful when cash is tight, but also quite a bit more work. https://plaintextaccounting.org/Budgeting has more on this topic. Using the budget report Historically this report has been confusing and fragile. hledger's version should be relatively robust and intuitive, but you may still find surprises. Here are more notes to help with learning and trou- bleshooting. o In the above example, expenses:bus and expenses:food are shown be- cause they have budget goals during the report period. o Their parent expenses is also shown, with budget goals aggregated from the children. o The subaccounts expenses:food:groceries and expenses:food:dining are not shown since they have no budget goal of their own, but they con- tribute to expenses:food's actual amount. o Unbudgeted accounts expenses:movies and expenses:gifts are also not shown, but they contribute to expenses's actual amount. o The other unbudgeted accounts income and assets:bank:checking are grouped as . o --depth or depth: can be used to limit report depth in the usual way (but will not reveal unbudgeted subaccounts). o Amounts are always inclusive of subaccounts (even in -l/--list mode). o Numbers displayed in a --budget report will not always agree with the totals, because of hidden unbudgeted accounts; this is normal. -E/--empty can be used to reveal the hidden accounts. o In the periodic rules used for setting budget goals, unbalanced post- ings are convenient. o You can filter budget reports with the usual queries, eg to focus on particular accounts. It's common to restrict them to just expenses. (The account is occasionally hard to exclude; this is because of date surprises, discussed below.) o When you have multiple currencies, you may want to convert them to one (-X COMM --infer-market-prices) and/or show just one at a time (cur:COMM). If you do need to show multiple currencies at once, --layout bare can be helpful. o You can "roll over" amounts (actual and budgeted) to the next period with --cumulative. See also: https://hledger.org/budgeting.html. Budget date surprises With small data, or when starting out, some of the generated budget goal transaction dates might fall outside the report periods. Eg with the following journal and report, the first period appears to have no expenses:food budget. (Also the account should be ex- cluded by the expenses query, but isn't.): ~ monthly in 2020 (expenses:food) $500 2020-01-15 expenses:food $400 assets:checking $ hledger bal --budget expenses Budget performance in 2020-01-15: || 2020-01-15 ===============++==================== || $400 expenses:food || 0 [ 0% of $500] ---------------++-------------------- || $400 [80% of $500] In this case, the budget goal transactions are generated on first days of of month (this can be seen with hledger print --forecast tag:gener- ated expenses). Whereas the report period defaults to just the 15th day of january (this can be seen from the report table's column head- ings). To fix this kind of thing, be more explicit about the report period (and/or the periodic rules' dates). In this case, adding -b 2020 does the trick. Selecting budget goals By default, the budget report uses all available periodic transaction rules to generate goals. This includes rules with a different report interval from your report. Eg if you have daily, weekly and monthly periodic rules, all of these will contribute to the goals in a monthly budget report. You can select a subset of periodic rules by providing an argument to the --budget flag. --budget=DESCPAT will match all periodic rules whose description contains DESCPAT, a case-insensitive substring (not a regular expression or query). This means you can give your periodic rules descriptions (remember that two spaces are needed between period expression and description), and then select from multiple budgets de- fined in your journal. Budgeting vs forecasting --forecast and --budget both use the periodic transaction rules in the journal to generate temporary transactions for reporting purposes. However they are separate features - though you can use both at the same time if you want. Here are some differences between them: --forecast --budget -------------------------------------------------------------------------- is a general option; it enables fore- is a balance command option; it casting with all reports selects the balance report's budget mode generates visible transactions which generates invisible transactions appear in reports which produce goal amounts generates forecast transactions from generates budget goal transac- after the last regular transaction, to tions throughout the report pe- the end of the report period; or with riod, optionally restricted by an argument --forecast=PERIODEXPR gen- periods specified in the peri- erates them throughout the specified odic transaction rules period, both optionally restricted by periods specified in the periodic transaction rules uses all periodic rules uses all periodic rules; or with an argument --budget=DESCPAT uses just the rules matched by DESCPAT Balance report layout The --layout option affects how balance and the other balance-like com- mands show multi-commodity amounts and commodity symbols. It can im- prove readability, for humans and/or machines (other software). It has four possible values: o --layout=wide[,WIDTH]: commodities are shown on a single line, op- tionally elided to WIDTH o --layout=tall: each commodity is shown on a separate line o --layout=bare: commodity symbols are in their own column, amounts are bare numbers o --layout=tidy: data is normalised to easily-consumed "tidy" form, with one row per data value. (This one is currently supported only by the balance command.) Here are the --layout modes supported by each output format Only CSV output supports all of them: - txt csv html json sql ------------------------------------- wide Y Y Y tall Y Y Y bare Y Y Y tidy Y Examples: Wide layout With many commodities, reports can be very wide: $ hledger -f examples/bcexample.hledger bal assets:us:etrade -3 -T -Y --layout=wide Balance changes in 2012-01-01..2014-12-31: || 2012 2013 2014 Total ==================++==================================================================================================================================================================================================================== Assets:US:ETrade || 10.00 ITOT, 337.18 USD, 12.00 VEA, 106.00 VHT 70.00 GLD, 18.00 ITOT, -98.12 USD, 10.00 VEA, 18.00 VHT -11.00 ITOT, 4881.44 USD, 14.00 VEA, 170.00 VHT 70.00 GLD, 17.00 ITOT, 5120.50 USD, 36.00 VEA, 294.00 VHT ------------------++-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- || 10.00 ITOT, 337.18 USD, 12.00 VEA, 106.00 VHT 70.00 GLD, 18.00 ITOT, -98.12 USD, 10.00 VEA, 18.00 VHT -11.00 ITOT, 4881.44 USD, 14.00 VEA, 170.00 VHT 70.00 GLD, 17.00 ITOT, 5120.50 USD, 36.00 VEA, 294.00 VHT A width limit reduces the width, but some commodities will be hidden: $ hledger -f examples/bcexample.hledger bal assets:us:etrade -3 -T -Y --layout=wide,32 Balance changes in 2012-01-01..2014-12-31: || 2012 2013 2014 Total ==================++=========================================================================================================================== Assets:US:ETrade || 10.00 ITOT, 337.18 USD, 2 more.. 70.00 GLD, 18.00 ITOT, 3 more.. -11.00 ITOT, 3 more.. 70.00 GLD, 17.00 ITOT, 3 more.. ------------------++--------------------------------------------------------------------------------------------------------------------------- || 10.00 ITOT, 337.18 USD, 2 more.. 70.00 GLD, 18.00 ITOT, 3 more.. -11.00 ITOT, 3 more.. 70.00 GLD, 17.00 ITOT, 3 more.. Tall layout Each commodity gets a new line (may be different in each column), and account names are repeated: $ hledger -f examples/bcexample.hledger bal assets:us:etrade -3 -T -Y --layout=tall Balance changes in 2012-01-01..2014-12-31: || 2012 2013 2014 Total ==================++================================================== Assets:US:ETrade || 10.00 ITOT 70.00 GLD -11.00 ITOT 70.00 GLD Assets:US:ETrade || 337.18 USD 18.00 ITOT 4881.44 USD 17.00 ITOT Assets:US:ETrade || 12.00 VEA -98.12 USD 14.00 VEA 5120.50 USD Assets:US:ETrade || 106.00 VHT 10.00 VEA 170.00 VHT 36.00 VEA Assets:US:ETrade || 18.00 VHT 294.00 VHT ------------------++-------------------------------------------------- || 10.00 ITOT 70.00 GLD -11.00 ITOT 70.00 GLD || 337.18 USD 18.00 ITOT 4881.44 USD 17.00 ITOT || 12.00 VEA -98.12 USD 14.00 VEA 5120.50 USD || 106.00 VHT 10.00 VEA 170.00 VHT 36.00 VEA || 18.00 VHT 294.00 VHT Bare layout Commodity symbols are kept in one column, each commodity has its own row, amounts are bare numbers, account names are repeated: $ hledger -f examples/bcexample.hledger bal assets:us:etrade -3 -T -Y --layout=bare Balance changes in 2012-01-01..2014-12-31: || Commodity 2012 2013 2014 Total ==================++============================================= Assets:US:ETrade || GLD 0 70.00 0 70.00 Assets:US:ETrade || ITOT 10.00 18.00 -11.00 17.00 Assets:US:ETrade || USD 337.18 -98.12 4881.44 5120.50 Assets:US:ETrade || VEA 12.00 10.00 14.00 36.00 Assets:US:ETrade || VHT 106.00 18.00 170.00 294.00 ------------------++--------------------------------------------- || GLD 0 70.00 0 70.00 || ITOT 10.00 18.00 -11.00 17.00 || USD 337.18 -98.12 4881.44 5120.50 || VEA 12.00 10.00 14.00 36.00 || VHT 106.00 18.00 170.00 294.00 Bare layout also affects CSV output, which is useful for producing data that is easier to consume, eg for making charts: $ hledger -f examples/bcexample.hledger bal assets:us:etrade -3 -O csv --layout=bare "account","commodity","balance" "Assets:US:ETrade","GLD","70.00" "Assets:US:ETrade","ITOT","17.00" "Assets:US:ETrade","USD","5120.50" "Assets:US:ETrade","VEA","36.00" "Assets:US:ETrade","VHT","294.00" "Total:","GLD","70.00" "Total:","ITOT","17.00" "Total:","USD","5120.50" "Total:","VEA","36.00" "Total:","VHT","294.00" Bare layout will sometimes display an extra row for the no-symbol com- modity, because of zero amounts (hledger treats zeroes as commod- ity-less, usually). This can break hledger-bar confusingly (workaround: add a cur: query to exclude the no-symbol row). Tidy layout This produces normalised "tidy data" (see https://cran.r-project.org/web/packages/tidyr/vignettes/tidy-data.html) where every variable has its own column and each row represents a sin- gle data point. This is the easiest kind of data for other software to consume: $ hledger -f examples/bcexample.hledger bal assets:us:etrade -3 -Y -O csv --layout=tidy "account","period","start_date","end_date","commodity","value" "Assets:US:ETrade","2012","2012-01-01","2012-12-31","GLD","0" "Assets:US:ETrade","2012","2012-01-01","2012-12-31","ITOT","10.00" "Assets:US:ETrade","2012","2012-01-01","2012-12-31","USD","337.18" "Assets:US:ETrade","2012","2012-01-01","2012-12-31","VEA","12.00" "Assets:US:ETrade","2012","2012-01-01","2012-12-31","VHT","106.00" "Assets:US:ETrade","2013","2013-01-01","2013-12-31","GLD","70.00" "Assets:US:ETrade","2013","2013-01-01","2013-12-31","ITOT","18.00" "Assets:US:ETrade","2013","2013-01-01","2013-12-31","USD","-98.12" "Assets:US:ETrade","2013","2013-01-01","2013-12-31","VEA","10.00" "Assets:US:ETrade","2013","2013-01-01","2013-12-31","VHT","18.00" "Assets:US:ETrade","2014","2014-01-01","2014-12-31","GLD","0" "Assets:US:ETrade","2014","2014-01-01","2014-12-31","ITOT","-11.00" "Assets:US:ETrade","2014","2014-01-01","2014-12-31","USD","4881.44" "Assets:US:ETrade","2014","2014-01-01","2014-12-31","VEA","14.00" "Assets:US:ETrade","2014","2014-01-01","2014-12-31","VHT","170.00" Balance report output As noted in Output format, if you choose HTML output (by using -O html or -o somefile.html), you can create a hledger.css file in the same di- rectory to customise the report's appearance. The HTML and FODS output formats can generate hyperlinks to a hledger-web register view for each account and period. E.g. if your hledger-web server is reachable at http://localhost:5000 then you might run the balance command with the extra option --base-url=http://local- host:5000. You can also produce relative links, like --base-url="some/path" or --base-url="".) Some useful balance reports Some frequently used balance options/reports are: o bal -M revenues expenses Show revenues/expenses in each month. Also available as the incomes- tatement command. o bal -M -H assets liabilities Show historical asset/liability balances at each month end. Also available as the balancesheet command. o bal -M -H assets liabilities equity Show historical asset/liability/equity balances at each month end. Also available as the balancesheetequity command. o bal -M assets not:receivable Show changes to liquid assets in each month. Also available as the cashflow command. Also: o bal -M expenses -2 -SA Show monthly expenses summarised to depth 2 and sorted by average amount. o bal -M --budget expenses Show monthly expenses and budget goals. o bal -M --valuechange investments Show monthly change in market value of investment assets. o bal investments --valuechange -D date:lastweek amt:'>1000' -STA [--invert] Show top gainers [or losers] last week roi Shows the time-weighted (TWR) and money-weighted (IRR) rate of return on your investments. Flags: --cashflow show all amounts that were used to compute returns --investment=QUERY query to select your investment transactions --profit-loss=QUERY --pnl query to select profit-and-loss or appreciation/valuation transactions At a minimum, you need to supply a query (which could be just an ac- count name) to select your investment(s) with --inv, and another query to identify your profit and loss transactions with --pnl. If you do not record changes in the value of your investment manually, or do not require computation of time-weighted return (TWR), --pnl could be an empty query (--pnl "" or --pnl STR where STR does not match any of your accounts). This command will compute and display the internalized rate of return (IRR, also known as money-weighted rate of return) and time-weighted rate of return (TWR) for your investments for the time period re- quested. IRR is always annualized due to the way it is computed, but TWR is reported both as a rate over the chosen reporting period and as an annual rate. Price directives will be taken into account if you supply appropriate --cost or --value flags (see VALUATION). Note, in some cases this report can fail, for these reasons: o Error (NotBracketed): No solution for Internal Rate of Return (IRR). Possible causes: IRR is huge (>1000000%), balance of investment be- comes negative at some point in time. o Error (SearchFailed): Failed to find solution for Internal Rate of Return (IRR). Either search does not converge to a solution, or con- verges too slowly. Examples: o Using roi to compute total return of investment in stocks: https://github.com/simonmichael/hledger/blob/master/examples/invest- ing/roi-unrealised.ledger o Cookbook > Return on Investment: https://hledger.org/roi.html Spaces and special characters in --inv and --pnl Note that --inv and --pnl's argument is a query, and queries could have several space-separated terms (see QUERIES). To indicate that all search terms form single command-line argument, you will need to put them in quotes (see Special characters): $ hledger roi --inv 'term1 term2 term3 ...' If any query terms contain spaces themselves, you will need an extra level of nested quoting, eg: $ hledger roi --inv="'Assets:Test 1'" --pnl="'Equity:Unrealized Profit and Loss'" Semantics of --inv and --pnl Query supplied to --inv has to match all transactions that are related to your investment. Transactions not matching --inv will be ignored. In these transactions, ROI will conside postings that match --inv to be "investment postings" and other postings (not matching --inv) will be sorted into two categories: "cash flow" and "profit and loss", as ROI needs to know which part of the investment value is your contributions and which is due to the return on investment. o "Cash flow" is depositing or withdrawing money, buying or selling as- sets, or otherwise converting between your investment commodity and any other commodity. Example: 2019-01-01 Investing in Snake Oil assets:cash -$100 investment:snake oil 2020-01-01 Selling my Snake Oil assets:cash $10 investment:snake oil = 0 o "Profit and loss" is change in the value of your investment: 2019-06-01 Snake Oil falls in value investment:snake oil = $57 equity:unrealized profit or loss All non-investment postings are assumed to be "cash flow", unless they match --pnl query. Changes in value of your investment due to "profit and loss" postings will be considered as part of your investment re- turn. Example: if you use --inv snake --pnl equity:unrealized, then postings in the example below would be classifed as: 2019-01-01 Snake Oil #1 assets:cash -$100 ; cash flow posting investment:snake oil ; investment posting 2019-03-01 Snake Oil #2 equity:unrealized pnl -$100 ; profit and loss posting snake oil ; investment posting 2019-07-01 Snake Oil #3 equity:unrealized pnl ; profit and loss posting cash -$100 ; cash flow posting snake oil $50 ; investment posting IRR and TWR explained "ROI" stands for "return on investment". Traditionally this was com- puted as a difference between current value of investment and its ini- tial value, expressed in percentage of the initial value. However, this approach is only practical in simple cases, where invest- ments receives no in-flows or out-flows of money, and where rate of growth is fixed over time. For more complex scenarios you need differ- ent ways to compute rate of return, and this command implements two of them: IRR and TWR. Internal rate of return, or "IRR" (also called "money-weighted rate of return") takes into account effects of in-flows and out-flows, and the time between them. Investment at a particular fixed interest rate is going to give you more interest than the same amount invested at the same interest rate, but made later in time. If you are withdrawing from your investment, your future gains would be smaller (in absolute numbers), and will be a smaller percentage of your initial investment, so your IRR will be smaller. And if you are adding to your investment, you will receive bigger absolute gains, which will be a bigger percent- age of your initial investment, so your IRR will be larger. As mentioned before, in-flows and out-flows would be any cash that you personally put in or withdraw, and for the "roi" command, these are the postings that match the query in the--inv argument and NOT match the query in the--pnl argument. If you manually record changes in the value of your investment as transactions that balance them against "profit and loss" (or "unreal- ized gains") account or use price directives, then in order for IRR to compute the precise effect of your in-flows and out-flows on the rate of return, you will need to record the value of your investement on or close to the days when in- or out-flows occur. In technical terms, IRR uses the same approach as computation of net present value, and tries to find a discount rate that makes net present value of all the cash flows of your investment to add up to zero. This could be hard to wrap your head around, especially if you haven't done discounted cash flow analysis before. Implementation of IRR in hledger should produce results that match the =XIRR formula in Excel. Second way to compute rate of return that roi command implements is called "time-weighted rate of return" or "TWR". Like IRR, it will ac- count for the effect of your in-flows and out-flows, but unlike IRR it will try to compute the true rate of return of the underlying asset, compensating for the effect that deposits and withdrawas have on the apparent rate of growth of your investment. TWR represents your investment as an imaginary "unit fund" where in-flows/ out-flows lead to buying or selling "units" of your invest- ment and changes in its value change the value of "investment unit". Change in "unit price" over the reporting period gives you rate of re- turn of your investment, and make TWR less sensitive than IRR to the effects of cash in-flows and out-flows. References: o Explanation of rate of return o Explanation of IRR o Explanation of TWR o IRR vs TWR o Examples of computing IRR and TWR and discussion of the limitations of both metrics Chart commands activity Show an ascii barchart of posting counts per interval. Flags: no command-specific flags The activity command displays an ascii histogram showing transaction counts by day, week, month or other reporting interval (by day is the default). With query arguments, it counts only matched transactions. Examples: $ hledger activity --quarterly 2008-01-01 ** 2008-04-01 ******* 2008-07-01 2008-10-01 ** Data generation commands close (equity) close prints several kinds of "closing" and/or "opening" transactions, useful in various situations: migrating balances to a new journal file, retaining earnings into equity, consolidating balances, viewing lot costs.. Like print, it prints valid journal entries. You can copy these into your journal file(s) when you are happy with how they look. Flags: --clopen[=TAGVAL] show closing and opening balances transactions, for AL accounts by default --close[=TAGVAL] show just a closing balances transaction --open[=TAGVAL] show just an opening balances transaction --assert[=TAGVAL] show a balance assertions transaction --assign[=TAGVAL] show a balance assignments transaction --retain[=TAGVAL] show a retain earnings transaction, for RX accounts by default -x --explicit show all amounts explicitly --show-costs show amounts with different costs separately --interleaved show source and destination postings together --assertion-type=TYPE =, ==, =* or ==* --close-desc=DESC set closing transaction's description --close-acct=ACCT set closing transaction's destination account --open-desc=DESC set opening transaction's description --open-acct=ACCT set opening transaction's source account --round=TYPE how much rounding or padding should be done when displaying amounts ? none - show original decimal digits, as in journal (default) soft - just add or remove decimal zeros to match precision hard - round posting amounts to precision (can unbalance transactions) all - also round cost amounts to precision (can unbalance transactions) close has six modes, selected by choosing one of the mode flags (--close is the default). They all do much the same operation, but with different defaults, useful in different situations. close --clopen This is useful if migrating balances to a new journal file at the start of a new year. It prints a "closing balances" transaction that zeroes out account balances (Asset and Liability accounts, by default), and an opposite "opening balances" transaction that restores them again. Typ- ically, you would run hledger close --clopen -e NEWYEAR >> $LEDGER_FILE and then move the opening transaction from the old file to the new file (and probably also update your LEDGER_FILE environment variable). Why might you do this ? If your reports are fast, you may not need it. But at some point you will probably want to partition your data by time, for performance or data integrity or regulatory reasons. A new file or set of files per year is common. Then, having each file/file- set "bookended" with opening and closing balance transactions will al- low you to freely pick and choose which files to read - just the cur- rent year, any past year, any sequence of years, or all of them - while showing correct account balances in each case. The earliest opening balances transaction sets correct starting balances, and any later closing/opening pairs will harmlessly cancel each other out. The balances will be transferred to and from equity:opening/closing balances by default. You can override this by using --close-acct and/or --open-acct. You can select a different set of accounts to close/open by providing an account query. Eg to add Equity accounts, provide arguments like assets liabilities equity or type:ALE. When migrating to a new file, you'll usually want to bring along the AL or ALE accounts, but not the RX accounts (Revenue, Expense). Assertions will be added indicating and checking the new balances of the closed/opened accounts. The generated transactions will have a clopen: tag. If the main jour- nal's base file name contains a number (eg a year number), the tag's value will be that base file name with the number incremented. Or you can choose the tag value yourself, by using --clopen=TAGVAL. close --close This prints just the closing balances transaction of --clopen. It is the default if you don't specify a mode. More customisation options are described below. Among other things, you can use close --close to generate a transaction moving the balances from any set of accounts, to a different account. (If you need to move just a portion of the balance, see hledger-move.) close --open This prints just the opening balances transaction of --clopen. (It is similar to Ledger's equity command.) close --assert This prints a transaction that asserts the account balances as they are on the end date (and adds an assert: tag). It could be useful as docu- mention and to guard against changes. close --assign This prints a transaction that assigns the account balances as they are on the end date (and adds an "assign:" tag). Unlike balance asser- tions, assignments will post changes to balances as needed to reach the specified amounts. This is another way to set starting balances when migrating to a new file, and it will set them correctly even in the presence of earlier files which do not have a closing balances transaction. However, it can hide errors, and disturb the accounting equation, so --clopen is usually recommended. close --retain This is like --close, but it closes Revenue and Expense account bal- ances by default. They will be transferred to equity:retained earn- ings, or another account specified with --close-acct. Revenues and expenses correspond to changes in equity. They are cate- gorised separately for reporting purposes, but traditionally at the end of each accounting period, businesses consolidate them into equity, This is called "retaining earnings", or "closing the books". In personal accounting, there's not much reason to do this, and most people don't. (One reason to do it is to help the balancesheetequity report show a zero total, demonstrating that the accounting equation (A-L=E) is satisfied.) close customisation In all modes, the following things can be overridden: o the accounts to be closed/opened, with account query arguments o the balancing account, with --close-acct=ACCT and/or --open-acct=ACCT o the transaction descriptions, with --close-desc=DESC and --open-desc=DESC o the transaction's tag value, with a --MODE=NEW option argument o the closing/opening dates, with -e OPENDATE By default, the closing date is yesterday, or the journal's end date, whichever is later; and the opening date is always one day after the closing date. You can change these by specifying a report end date; the closing date will be the last day of the report period. Eg -e 2024 means "close on 2023-12-31, open on 2024-01-01". With --x/--explicit, the balancing amount will be shown explicitly, and if it involves multiple commodities, a separate posting will be gener- ated for each of them (similar to print -x). With --interleaved, each individual transfer is shown with source and destination postings next to each other (perhaps useful for trou- bleshooting). With --show-costs, balances' costs are also shown, with different costs kept separate. This may generate very large journal entries, if you have many currency conversions or investment transactions. close --show-costs is currently the best way to view investment lots with hledger. (To move or dispose of lots, see the more capable hledger-move script.) close and balance assertions close adds balance assertions verifying that the accounts have been re- set to zero in a closing transaction or restored to their previous bal- ances in an opening transaction. These provide useful error checking, but you can ignore them temporarily with -I, or remove them if you pre- fer. Single-commodity, subaccount-exclusive balance assertions (=) are gen- erated by default. This can be changed with --assertion-type='==*' (eg). When running close you should probably avoid using -C, -R, status: (filtering by status or realness) or --auto (generating postings), since the generated balance assertions would then require these. Transactions with multiple dates (eg posting dates) spanning the file boundary also can disrupt the balance assertions: 2023-12-30 a purchase made in december, cleared in january expenses:food 5 assets:bank:checking -5 ; date: 2023-01-02 To solve this you can transfer the money to and from a temporary ac- count, splitting the multi-day transaction into two single-day transac- tions: ; in 2022.journal: 2022-12-30 a purchase made in december, cleared in january expenses:food 5 equity:pending -5 ; in 2023.journal: 2023-01-02 last year's transaction cleared equity:pending 5 = 0 assets:bank:checking -5 close examples Retain earnings Record 2022's revenues/expenses as retained earnings on 2022-12-31, ap- pending the generated transaction to the journal: $ hledger close --retain -f 2022.journal -p 2022 >> 2022.journal After this, to see 2022's revenues and expenses you must exclude the retain earnings transaction: $ hledger -f 2022.journal is not:desc:'retain earnings' Migrate balances to a new file Close assets/liabilities on 2022-12-31 and re-open them on 2023-01-01: $ hledger close --clopen -f 2022.journal -p 2022 # copy/paste the closing transaction to the end of 2022.journal # copy/paste the opening transaction to the start of 2023.journal After this, to see 2022's end-of-year balances you must exclude the closing balances transaction: $ hledger -f 2022.journal bs not:desc:'closing balances' For more flexibility, it helps to tag closing and opening transactions with eg clopen:NEWYEAR, then you can ensure correct balances by exclud- ing all opening/closing transactions except the first, like so: $ hledger bs -Y -f 2021.j -f 2022.j -f 2023.j expr:'tag:clopen=2021 or not tag:clopen' $ hledger bs -Y -f 2021.j -f 2022.j expr:'tag:clopen=2021 or not tag:clopen' $ hledger bs -Y -f 2022.j -f 2023.j expr:'tag:clopen=2022 or not tag:clopen' $ hledger bs -Y -f 2021.j expr:'tag:clopen=2021 or not tag:clopen' $ hledger bs -Y -f 2022.j expr:'tag:clopen=2022 or not tag:clopen' $ hledger bs -Y -f 2023.j # unclosed file, no query needed More detailed close examples See examples/multi-year. rewrite Print all transactions, rewriting the postings of matched transactions. For now the only rewrite available is adding new postings, like print --auto. Flags: --add-posting='ACCT AMTEXPR' add a posting to ACCT, which may be parenthesised. AMTEXPR is either a literal amount, or *N which means the transaction's first matched amount multiplied by N (a decimal number). Two spaces separate ACCT and AMTEXPR. --diff generate diff suitable as an input for patch tool This is a start at a generic rewriter of transaction entries. It reads the default journal and prints the transactions, like print, but adds one or more specified postings to any transactions matching QUERY. The posting amounts can be fixed, or a multiplier of the existing transac- tion's first posting amount. Examples: $ hledger-rewrite.hs ^income --add-posting '(liabilities:tax) *.33 ; income tax' --add-posting '(reserve:gifts) $100' $ hledger-rewrite.hs expenses:gifts --add-posting '(reserve:gifts) *-1"' $ hledger-rewrite.hs -f rewrites.hledger rewrites.hledger may consist of entries like: = ^income amt:<0 date:2017 (liabilities:tax) *0.33 ; tax on income (reserve:grocery) *0.25 ; reserve 25% for grocery (reserve:) *0.25 ; reserve 25% for grocery Note the single quotes to protect the dollar sign from bash, and the two spaces between account and amount. More: $ hledger rewrite [QUERY] --add-posting "ACCT AMTEXPR" ... $ hledger rewrite ^income --add-posting '(liabilities:tax) *.33' $ hledger rewrite expenses:gifts --add-posting '(budget:gifts) *-1"' $ hledger rewrite ^income --add-posting '(budget:foreign currency) *0.25 JPY; diversify' Argument for --add-posting option is a usual posting of transaction with an exception for amount specification. More precisely, you can use '*' (star symbol) before the amount to indicate that that this is a factor for an amount of original matched posting. If the amount in- cludes a commodity name, the new posting amount will be in the new com- modity; otherwise, it will be in the matched posting amount's commod- ity. Re-write rules in a file During the run this tool will execute so called "Automated Transac- tions" found in any journal it process. I.e instead of specifying this operations in command line you can put them in a journal file. $ rewrite-rules.journal Make contents look like this: = ^income (liabilities:tax) *.33 = expenses:gifts budget:gifts *-1 assets:budget *1 Note that '=' (equality symbol) that is used instead of date in trans- actions you usually write. It indicates the query by which you want to match the posting to add new ones. $ hledger rewrite -f input.journal -f rewrite-rules.journal > rewritten-tidy-output.journal This is something similar to the commands pipeline: $ hledger rewrite -f input.journal '^income' --add-posting '(liabilities:tax) *.33' \ | hledger rewrite -f - expenses:gifts --add-posting 'budget:gifts *-1' \ --add-posting 'assets:budget *1' \ > rewritten-tidy-output.journal It is important to understand that relative order of such entries in journal is important. You can re-use result of previously added post- ings. Diff output format To use this tool for batch modification of your journal files you may find useful output in form of unified diff. $ hledger rewrite --diff -f examples/sample.journal '^income' --add-posting '(liabilities:tax) *.33' Output might look like: --- /tmp/examples/sample.journal +++ /tmp/examples/sample.journal @@ -18,3 +18,4 @@ 2008/01/01 income - assets:bank:checking $1 + assets:bank:checking $1 income:salary + (liabilities:tax) 0 @@ -22,3 +23,4 @@ 2008/06/01 gift - assets:bank:checking $1 + assets:bank:checking $1 income:gifts + (liabilities:tax) 0 If you'll pass this through patch tool you'll get transactions contain- ing the posting that matches your query be updated. Note that multiple files might be update according to list of input files specified via --file options and include directives inside of these files. Be careful. Whole transaction being re-formatted in a style of output from hledger print. See also: https://github.com/simonmichael/hledger/issues/99 rewrite vs. print --auto This command predates print --auto, and currently does much the same thing, but with these differences: o with multiple files, rewrite lets rules in any file affect all other files. print --auto uses standard directive scoping; rules affect only child files. o rewrite's query limits which transactions can be rewritten; all are printed. print --auto's query limits which transactions are printed. o rewrite applies rules specified on command line or in the journal. print --auto applies rules specified in the journal. Maintenance commands check Check for various kinds of errors in your data. Flags: no command-specific flags hledger provides a number of built-in correctness checks to help vali- date your data and prevent errors. Some are run automatically, some when you enable --strict mode; or you can run any of them on demand by providing them as arguments to the check command. check produces no output and a zero exit code if all is well. Eg: hledger check # run basic checks hledger check -s # run basic and strict checks hledger check ordereddates payees # run basic checks and two others If you are an Emacs user, you can also configure flycheck-hledger to run these checks, providing instant feedback as you edit the journal. Here are the checks currently available. Generally, they are performed in the order they are shown here (and only the first failure is re- ported). Basic checks These important checks are performed by default, by almost all hledger commands: o parseable - data files are in a supported format, with no syntax er- rors and no invalid include directives. This ensures that all files exist and are readable. o autobalanced - all transactions are balanced, after automatically in- ferring missing amounts and conversion rates and then converting amounts to cost. This ensures that each transaction's entry is well formed. o assertions - all balance assertions in the journal are passing. Bal- ance assertions are a strong defense against errors; they help catch many problems. If this check gets in your way, you can disable it with -I/--ignore-assertions. Or you can add that to your config file to disable it by default (and then use -s/--strict or hledger check assertions to enable it). Strict checks These additional checks are performed by all commands when the -s/--strict flag is used (strict mode). They provide extra er- ror-catching power to keep your data clean and correct. Strict mode also always enables the assertions check. o balanced - like autobalanced, but all conversions between commodities must use explicit cost notation or equity postings. This prevents wrong conversions caused by typos. o commodities - all commodity symbols used must be declared. This guards against mistyping or omitting commodity symbols. Declaring commodities also sets their precision for display and transaction balancing. o accounts - all account names used must be declared. This prevents the use of mis-spelled or outdated account names. Other checks These are not wanted by everyone, but can be run using the check com- mand: o tags - all tags used must be declared. This prevents mis-spelled tag names. Note hledger fairly often finds unintended tags in comments. o payees - all payees used in transactions must be declared. This will force you to declare any new payee name before using it. Most people will probably find this a bit too strict. o ordereddates - within each file, transactions must be ordered by date. This is a simple and effective error catcher. It's not in- cluded in strict mode, but you can add it by running hledger check -s ordereddates. If enabled, this check is performed before balance as- sertions. o recentassertions - all accounts with balance assertions must have one that's within the 7 days before their latest posting. This will en- courage adding balance assertions for your active asset/liability ac- counts, which in turn should encourage you to reconcile regularly with those real world balances - another strong defense against er- rors. hledger close --assert can help generate assertion entries. Over time the older assertions become somewhat redundant, and you can remove them if you like (they don't affect performance much, but they add some noise to the journal). o uniqueleafnames - no two accounts may have the same last account name part (eg the checking in assets:bank:checking). This ensures each account can be matched by a unique short name, easier to remember and to type. Custom checks You can build your own custom checks with add-on command scripts. See also Cookbook > Scripting. Here are some examples from hledger/bin/: o hledger-check-tagfiles - all tag values containing / (a forward slash) exist as file paths o hledger-check-fancyassertions - more complex balance assertions are passing diff Compares a particular account's transactions in two input files. It shows any transactions to this account which are in one file but not in the other. Flags: no command-specific flags More precisely: for each posting affecting this account in either file, this command looks for a corresponding posting in the other file which posts the same amount to the same account (ignoring date, description, etc). Since it compares postings, not transactions, this also works when mul- tiple bank transactions have been combined into a single journal entry. This command is useful eg if you have downloaded an account's transac- tions from your bank (eg as CSV data): when hledger and your bank dis- agree about the account balance, you can compare the bank data with your journal to find out the cause. Examples: $ hledger diff -f $LEDGER_FILE -f bank.csv assets:bank:giro These transactions are in the first file only: 2014/01/01 Opening Balances assets:bank:giro EUR ... ... equity:opening balances EUR -... These transactions are in the second file only: setup Check the status of the hledger installation. Flags: no command-specific flags setup tests your hledger installation and prints a list of results, sometimes with helpful hints. This is a good first command to run af- ter installing hledger. Also after upgrading, or when something's not working, or just when you want a reminder of where things are. It makes one network request to detect the latest hledger release ver- sion. It's ok if this fails or times out. It will use ANSI color by default, unless disabled by NO_COLOR or --color=n. It does not use a pager or a config file. It expects that the hledger version you are running is installed in your PATH. If not, it will stop until you have done that (to keep things simple). Example: $ hledger setup Checking your hledger setup.. Legend: good, neutral, unknown, warning hledger * is a released version ? no hledger 1.42.99-gbca4b39c5-20250425, mac-aarch64 * is up to date ? yes 1.42.99 installed, latest is 1.42.1 * is a native binary for this machine ? yes aarch64 * is installed in PATH ? yes /Users/simon/.local/bin/hledger * has a system text encoding configured ? yes UTF-8, data files should use this encoding * has a user config file ? (optional) no * current directory has a local config ? yes /Users/simon/src/hledger/hledger.conf * the config file is readable ? yes /Users/simon/src/hledger/hledger.conf terminal * the NO_COLOR variable is defined ? no * --color is configured by config file ? no * hledger will use color by default ? yes * the PAGER variable is defined ? yes less * --pager is configured by config file ? no * hledger will use a pager when needed ? yes /opt/homebrew/bin/less * the LESS variable is defined ? yes * the HLEDGER_LESS variable is defined ? no * adjusting LESS variable for color etc. ? yes * --pretty is enabled by config file ? no tables will use ASCII characters * bash shell completions are installed ? ? * zsh shell completions are installed ? ? journal * the LEDGER_FILE variable is defined ? yes /Users/simon/finance/2025/2025.journal * a default journal file is readable ? yes /Users/simon/finance/2025/2025.journal * it includes additional files ? yes 15 * all commodities are declared ? yes 10 * all accounts are declared ? yes 160 * all accounts have types ? no 14 untyped * accounts of each type were detected ? yes ALERXCV * commodities/accounts are checked ? no use -s to check commodities/accounts * balance assertions are checked ? yes use -I to ignore assertions test Run built-in unit tests. Flags: no command-specific flags This command runs the unit tests built in to hledger and hledger-lib, printing the results on stdout. If any test fails, the exit code will be non-zero. This is mainly used by hledger developers, but you can also use it to sanity-check the installed hledger executable on your platform. All tests are expected to pass - if you ever see a failure, please report as a bug! Any arguments before a -- argument will be passed to the tasty test runner as test-selecting -p patterns, and any arguments after -- will be passed to tasty unchanged. Examples: $ hledger test # run all unit tests $ hledger test balance # run tests with "balance" in their name $ hledger test -- -h # show tasty's options PART 5: COMMON TASKS Here are some quick examples of how to do some basic tasks with hledger. Getting help Here's how to list commands and view options and command docs: $ hledger # show available commands $ hledger --help # show common options $ hledger CMD --help # show CMD's options, common options and CMD's documentation You can also view your hledger version's manual in several formats by using the help command. Eg: $ hledger help # show the hledger manual with info, man or $PAGER (best available) $ hledger help journal # show the journal topic in the hledger manual $ hledger help --help # find out more about the help command To view manuals and introductory docs on the web, visit https://hledger.org. Chat and mail list support and discussion archives can be found at https://hledger.org/support. Constructing command lines hledger has a flexible command line interface. We strive to keep it simple and ergonomic, but if you run into one of the sharp edges de- scribed in OPTIONS, here are some tips that might help: o command-specific options must go after the command (it's fine to put common options there too: hledger CMD OPTS ARGS) o you can run addon commands via hledger (hledger ui [ARGS]) or di- rectly (hledger-ui [ARGS]) o enclose "problematic" arguments in single quotes o if needed, also add a backslash to hide regular expression metachar- acters from the shell o to see how a misbehaving command line is being parsed, add --debug=2. Starting a journal file hledger looks for your accounting data in a journal file, $HOME/.hledger.journal by default: $ hledger stats The hledger journal file "/Users/simon/.hledger.journal" was not found. Please create it first, eg with "hledger add" or a text editor. Or, specify an existing journal file with -f or LEDGER_FILE. You can override this by setting the LEDGER_FILE environment variable (see below). It's a good practice to keep this important file under version control, and to start a new file each year. So you could do something like this: $ mkdir ~/finance $ cd ~/finance $ git init Initialized empty Git repository in /Users/simon/finance/.git/ $ touch 2023.journal $ echo "export LEDGER_FILE=$HOME/finance/2023.journal" >> ~/.profile $ source ~/.profile $ hledger stats Main file : /Users/simon/finance/2023.journal Included files : Transactions span : to (0 days) Last transaction : none Transactions : 0 (0.0 per day) Transactions last 30 days: 0 (0.0 per day) Transactions last 7 days : 0 (0.0 per day) Payees/descriptions : 0 Accounts : 0 (depth 0) Commodities : 0 () Market prices : 0 () Setting LEDGER_FILE How to set LEDGER_FILE permanently depends on your setup: On unix and mac, running these commands in the terminal will work for many people; adapt as needed: $ echo 'export LEDGER_FILE=~/finance/2023.journal' >> ~/.profile $ source ~/.profile When correctly configured, in a new terminal window env | grep LEDGER_FILE will show your file, and so will hledger files. On mac, this additional step might be helpful for GUI applications (like Emacs started from the dock): add an entry to ~/.MacOSX/environ- ment.plist like { "LEDGER_FILE" : "~/finance/2023.journal" } and then run killall Dock in a terminal window (or restart the ma- chine). On Windows, see https://www.java.com/en/download/help/path.html, or try running these commands in a powershell window (let us know if it per- sists across a reboot, and if you need to be an Administrator): > CD > MKDIR finance > SETX LEDGER_FILE "C:\Users\USERNAME\finance\2023.journal" When correctly configured, in a new terminal window $env:LEDGER_FILE will show the file path, and so will hledger files. Setting opening balances Pick a starting date for which you can look up the balances of some real-world assets (bank accounts, wallet..) and liabilities (credit cards..). To avoid a lot of data entry, you may want to start with just one or two accounts, like your checking account or cash wallet; and pick a re- cent starting date, like today or the start of the week. You can al- ways come back later and add more accounts and older transactions, eg going back to january 1st. Add an opening balances transaction to the journal, declaring the bal- ances on this date. Here are two ways to do it: o The first way: open the journal in any text editor and save an entry like this: 2023-01-01 * opening balances assets:bank:checking $1000 = $1000 assets:bank:savings $2000 = $2000 assets:cash $100 = $100 liabilities:creditcard $-50 = $-50 equity:opening/closing balances These are start-of-day balances, ie whatever was in the account at the end of the previous day. The * after the date is an optional status flag. Here it means "cleared & confirmed". The currency symbols are optional, but usually a good idea as you'll be dealing with multiple currencies sooner or later. The = amounts are optional balance assertions, providing extra error checking. o The second way: run hledger add and follow the prompts to record a similar transaction: $ hledger add Adding transactions to journal file /Users/simon/finance/2023.journal Any command line arguments will be used as defaults. Use tab key to complete, readline keys to edit, enter to accept defaults. An optional (CODE) may follow transaction dates. An optional ; COMMENT may follow descriptions or amounts. If you make a mistake, enter < at any prompt to go one step backward. To end a transaction, enter . when prompted. To quit, enter . at a date prompt or press control-d or control-c. Date [2023-02-07]: 2023-01-01 Description: * opening balances Account 1: assets:bank:checking Amount 1: $1000 Account 2: assets:bank:savings Amount 2 [$-1000]: $2000 Account 3: assets:cash Amount 3 [$-3000]: $100 Account 4: liabilities:creditcard Amount 4 [$-3100]: $-50 Account 5: equity:opening/closing balances Amount 5 [$-3050]: Account 6 (or . or enter to finish this transaction): . 2023-01-01 * opening balances assets:bank:checking $1000 assets:bank:savings $2000 assets:cash $100 liabilities:creditcard $-50 equity:opening/closing balances $-3050 Save this transaction to the journal ? [y]: Saved. Starting the next transaction (. or ctrl-D/ctrl-C to quit) Date [2023-01-01]: . If you're using version control, this could be a good time to commit the journal. Eg: $ git commit -m 'initial balances' 2023.journal Recording transactions As you spend or receive money, you can record these transactions using one of the methods above (text editor, hledger add) or by using the hledger-iadd or hledger-web add-ons, or by using the import command to convert CSV data downloaded from your bank. Here are some simple transactions, see the hledger_journal(5) manual and hledger.org for more ideas: 2023/1/10 * gift received assets:cash $20 income:gifts 2023.1.12 * farmers market expenses:food $13 assets:cash 2023-01-15 paycheck income:salary assets:bank:checking $1000 Reconciling Periodically you should reconcile - compare your hledger-reported bal- ances against external sources of truth, like bank statements or your bank's website - to be sure that your ledger accurately represents the real-world balances (and, that the real-world institutions have not made a mistake!). This gets easy and fast with (1) practice and (2) frequency. If you do it daily, it can take 2-10 minutes. If you let it pile up, expect it to take longer as you hunt down errors and dis- crepancies. A typical workflow: 1. Reconcile cash. Count what's in your wallet. Compare with what hledger reports (hledger bal cash). If they are different, try to remember the missing transaction, or look for the error in the al- ready-recorded transactions. A register report can be helpful (hledger reg cash). If you can't find the error, add an adjustment transaction. Eg if you have $105 after the above, and can't explain the missing $2, it could be: 2023-01-16 * adjust cash assets:cash $-2 = $105 expenses:misc 2. Reconcile checking. Log in to your bank's website. Compare today's (cleared) balance with hledger's cleared balance (hledger bal check- ing -C). If they are different, track down the error or record the missing transaction(s) or add an adjustment transaction, similar to the above. Unlike the cash case, you can usually compare the trans- action history and running balance from your bank with the one re- ported by hledger reg checking -C. This will be easier if you gen- erally record transaction dates quite similar to your bank's clear- ing dates. 3. Repeat for other asset/liability accounts. Tip: instead of the register command, use hledger-ui to see a live-up- dating register while you edit the journal: hledger-ui --watch --regis- ter checking -C After reconciling, it could be a good time to mark the reconciled transactions' status as "cleared and confirmed", if you want to track that, by adding the * marker. Eg in the paycheck transaction above, insert * between 2023-01-15 and paycheck If you're using version control, this can be another good time to com- mit: $ git commit -m 'txns' 2023.journal Reporting Here are some basic reports. Show all transactions: $ hledger print 2023-01-01 * opening balances assets:bank:checking $1000 assets:bank:savings $2000 assets:cash $100 liabilities:creditcard $-50 equity:opening/closing balances $-3050 2023-01-10 * gift received assets:cash $20 income:gifts 2023-01-12 * farmers market expenses:food $13 assets:cash 2023-01-15 * paycheck income:salary assets:bank:checking $1000 2023-01-16 * adjust cash assets:cash $-2 = $105 expenses:misc Show account names, and their hierarchy: $ hledger accounts --tree assets bank checking savings cash equity opening/closing balances expenses food misc income gifts salary liabilities creditcard Show all account totals: $ hledger balance $4105 assets $4000 bank $2000 checking $2000 savings $105 cash $-3050 equity:opening/closing balances $15 expenses $13 food $2 misc $-1020 income $-20 gifts $-1000 salary $-50 liabilities:creditcard -------------------- 0 Show only asset and liability balances, as a flat list, limited to depth 2: $ hledger bal assets liabilities -2 $4000 assets:bank $105 assets:cash $-50 liabilities:creditcard -------------------- $4055 Show the same thing without negative numbers, formatted as a simple balance sheet: $ hledger bs -2 Balance Sheet 2023-01-16 || 2023-01-16 ========================++============ Assets || ------------------------++------------ assets:bank || $4000 assets:cash || $105 ------------------------++------------ || $4105 ========================++============ Liabilities || ------------------------++------------ liabilities:creditcard || $50 ------------------------++------------ || $50 ========================++============ Net: || $4055 The final total is your "net worth" on the end date. (Or use bse for a full balance sheet with equity.) Show income and expense totals, formatted as an income statement: hledger is Income Statement 2023-01-01-2023-01-16 || 2023-01-01-2023-01-16 ===============++======================= Revenues || ---------------++----------------------- income:gifts || $20 income:salary || $1000 ---------------++----------------------- || $1020 ===============++======================= Expenses || ---------------++----------------------- expenses:food || $13 expenses:misc || $2 ---------------++----------------------- || $15 ===============++======================= Net: || $1005 The final total is your net income during this period. Show transactions affecting your wallet, with running total: $ hledger register cash 2023-01-01 opening balances assets:cash $100 $100 2023-01-10 gift received assets:cash $20 $120 2023-01-12 farmers market assets:cash $-13 $107 2023-01-16 adjust cash assets:cash $-2 $105 Show weekly posting counts as a bar chart: $ hledger activity -W 2019-12-30 ***** 2023-01-06 **** 2023-01-13 **** Migrating to a new file At the end of the year, you may want to continue your journal in a new file, so that old transactions don't slow down or clutter your reports, and to help ensure the integrity of your accounting history. See the close command. If using version control, don't forget to git add the new file. BUGS We welcome bug reports in the hledger issue tracker (https://bugs.hledger.org), or on the hledger chat or mail list (https://hledger.org/support). Some known issues and limitations: A system locale with a suitable text encoding must be configured to work with non-ascii data. (See Text encoding, Troubleshooting.) On Microsoft Windows, depending what kind of terminal window you use, non-ascii characters, ANSI text formatting, and/or the add command's TAB key for completion, may not be supported. When processing large data files, hledger uses more memory than Ledger. Troubleshooting Here are some common issues you might encounter when you run hledger, and how to resolve them (and remember also you can usually get quick Support): PATH issues: I get an error like "No command 'hledger' found" Depending how you installed hledger, the executables may not be in your shell's PATH. Eg on unix systems, stack installs hledger in ~/.lo- cal/bin and cabal installs it in ~/.cabal/bin. You may need to add one of these directories to your shell's PATH, and/or open a new terminal window. LEDGER_FILE issues: I configured LEDGER_FILE but hledger is not using it o LEDGER_FILE should be a real environment variable, not just a shell variable. Eg on unix, the command env | grep LEDGER_FILE should show it. You may need to use export (see https://stackover- flow.com/a/7411509). On Windows, $env:LEDGER_FILE should show it. o You may need to force your shell to see the new configuration. A simple way is to close your terminal window and open a new one. Text decoding issues: I get errors like "Illegal byte sequence" or "In- valid or incomplete multibyte or wide character" or "commitAndRelease- Buffer: invalid argument (invalid character)" hledger usually needs its input to be decodable with the system lo- cale's text encoding. See Text encoding and Install: Text encoding. COMPATIBILITY ISSUES: hledger gives an error with my Ledger file Not all of Ledger's journal file syntax or feature set is supported. See hledger and Ledger for full details. AUTHORS Simon Michael and contributors. See http://hledger.org/CREDITS.html COPYRIGHT Copyright 2007-2023 Simon Michael and contributors. LICENSE Released under GNU GPL v3 or later. SEE ALSO hledger(1), hledger-ui(1), hledger-web(1), ledger(1) hledger-1.50 September 2025 HLEDGER(1)