!"#$%&'()*+,-./012345678 9 : ; < = > ? @ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z [ \ ] ^ _ ` a b c d e f g h i j k l m n o pqrstuvwxyz{|}~      ! " # $ % & ' ( ) * + , - . / 0 1 2 3 4 5 6 7 8 9 : ; < = > ? @ A B C D E F G HIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{!|!}!~!!!!!!!!!!!!!!!!!!!""""#############$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$%%%%%&&&&&&&''''''''''''''(((())))))))))))))******+++++++++++++,,--../portable provisional$John Goerzen <jgoerzen@complete.org>None 2A type to standardize some common uses of GetOpt. AThe first component of the tuple is the long name of the option. MThe second component is empty if there is no arg, or has the arg otherwise. ?Simple command line parser -- a basic wrapper around the system's Odefault getOpt. See the System.Console.GetOpt manual for a description of the first two parameters. 3The third parameter is a usage information header. DThe return value consists of the list of parsed flags and a list of non-option arguments.  Similar to  2, but takes an additional function that validates Hthe post-parse command-line arguments. This is useful, for example, in Nsituations where there are two arguments that are mutually-exclusive and only )one may legitimately be given at a time. IThe return value of the function indicates whether or not it detected an Kerror condition. If it returns Nothing, there is no error. If it returns :Just String, there was an error, described by the String. Handle a required argument. Handle an optional argument.   Name of arg  Name of arg    portable provisional$John Goerzen <jgoerzen@complete.org>NoneIThe type representing a Debian version number. This type is an instance of , but you can also use  if you prefer. =The type representing the contents of a Debian control file, Ior any control-like file (such as the output from apt-cache show, etc.) 'Compare the versions of two packages.   Version 1  Operator  Version 2    /portable provisional%John Goerzen <jgoerzen@complete.org> Noneportable provisional%John Goerzen <jgoerzen@complete.org> Noneportable provisional$John Goerzen <jgoerzen@complete.org>None !"#$%&'()*+,-./ !"#$%&'()*+,-./ !"#$%&'()/.-,+*   !"#$%&'()*+,-./portable provisional%John Goerzen <jgoerzen@complete.org> None0/The name of the null device. NUL: on Windows, /dev/null everywhere else. 0    !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~000 Ian Lynagh, None4Returns (Data, Remainder) 5123456123456346215+123456None 777 7 portable provisionalIan Lynagh <igloo@earth.li>None8IAnything we want to work out the MD5 of must be an instance of class MD5 ECThe simplest function, gives you the MD5 of a string as 4-tuple of  32bit words. F.Returns a hex number ala the md5sum program. G1Returns an integer equivalent to hex number from F. '89:;<=>?@ABCDEFG89:;<=>?@ABCDEFGEFG89:;BCD@A>?<= 89:;<=>?@ABCDEFG portable provisional%John Goerzen <jgoerzen@complete.org> NoneHIJKHIJKHIJKHIJK portable provisional%John Goerzen <jgoerzen@complete.org> NoneLMNOLMNOLMNOLMNO $portable to platforms with rawSystem provisional%John Goerzen <jgoerzen@complete.org> NonePAReturns a list representing the bytes that comprise a data type.  Example: 8 getBytes (0x12345678::Int) -> [0x12, 0x34, 0x56, 0x78] QThe opposite of P), this function builds a number based on its component bytes. FResults are undefined if any components of the input list are > 0xff! RConverts a Char to a Word8. SConverts a String to a [Word8]. TConverts a Word8 to a Char. UConverts a [Word8] to a String. PQRSTUPQRSTUPQRSTUPQRSTU portable provisional$John Goerzen <jgoerzen@complete.org>NoneV<A section represents a compressed component in a GZip file. #Every GZip file has at least one. W,Stored on-disk at the end of each section. Y,The size of the original, decompressed data Z:The stored GZip CRC-32 of the original, decompressed data [KWhether or not the stored CRC-32 matches the calculated CRC-32 of the data \>The data structure representing the GZip header. This occurs at the beginning of each V on disk. ^3Compression method. Only 8 is defined at present. c'Modification time of the original file d Extra flags eCreating operating system gOther problem arose h8Compressed with something other than method 8 (deflate) iCouldn't find a GZip header jCRC-32 check failed First two bytes of file Flags Flags Flags Flags k4Read a GZip file, decompressing all sections found. @Writes the decompressed data stream to the given output handle. IReturns Nothing if the action was successful, or Just GZipError if there Gwas a problem. If there was a problem, the data written to the output handle should be discarded. l=Read a GZip file, decompressing all sections that are found. GReturns a decompresed data stream and Nothing, or an unreliable string Fand Just (error). If you get anything other than Nothing, the String returned should be discarded. mRead all sections. n5Read one section, returning (ThisSection, Remainder)  Read the file's compressed data, returning - (Decompressed, Calculated CRC32, Remainder) o3Read the GZip header. Return (Header, Remainder). #VWXYZ[\]^_`abcdefghijk Input handle Output handle lmnoVWXYZ[\]^_`abcdefghijklmno\]^_`abcdeVfjihgWXYZ[lkmonVWXYZ[\ ]^_`abcdefjihgklmnoportable provisional%John Goerzen <jgoerzen@complete.org> Nonep-Take the first item out of a 3 element tuple q.Take the second item out of a 3 element tuple r-Take the third item out of a 3 element tuple pqrpqrpqrpqrportable provisional%John Goerzen <jgoerzen@complete.org> Nones@Pulls a Just value out of a Maybe value. If the Maybe value is *Nothing, raises an exception with error. tLike s5, but lets you customize the error message raised if Nothing is supplied. ststststportable provisional%John Goerzen <jgoerzen@complete.org> NoneuHConverts a Maybe value to an Either value, using the supplied parameter +as the Left value if the Maybe is Nothing. %This function can be interpreted as:  +maybeToEither :: e -> Maybe a -> Either e aYIts definition is given as it is so that it can be used in the Error and related monads. vPulls a Right7 value out of an Either value. If the Either value is Left, raises an exception with error. wLike v4, but can raise a specific message with the error. xETakes an either and transforms it into something of the more generic MonadError class. y+Take a Left to a value, crashes on a Right z+Take a Right to a value, crashes on a Left {/Take an Either, and return the value inside it u8(Left e) will be returned if the Maybe value is Nothing /(Right a) will be returned if this is (Just a) vwxyz{uvwxyz{uvwxyz{uvwxyz{systems with networking provisional$John Goerzen <jgoerzen@complete.org>None|<Sets up the system for networking. Similar to the built-in BwithSocketsDo (and actually, calls it), but also sets the SIGPIPE #handler so that signal is ignored.  Example: # main = niceSocketsDo $ do { ... } |}~|}~|}~|}~systems with networking experimental%John Goerzen <jgoerzen@complete.org> None The main handler type. *The first parameter is the socket itself. 2The second is the address of the remote endpoint. 0The third is the address of the local endpoint. Options for your server. 7Get Default options. You can always modify it later. #Takes some options and sets up the . I will bind Band begin listening, but will not accept any connections itself. 4Close the socket server. Does not terminate active handlers, if any. +Handle one incoming request from the given . ,Handle all incoming requests from the given . 0Convenience function to completely set up a TCP # and handle all incoming requests. !This function is literally this: serveTCPforever options func = - do sockserv <- setupSocketServer options " serveForever sockserv func 4Log each incoming connection using the interface in System.Log.Logger. .Log when the incoming connection disconnects. =Also, log any failures that may occur in the child handler. 5Handle each incoming connection in its own thread to  make the server multi-tasking. 9Give your handler function a Handle instead of a Socket. JThe Handle will be opened with ReadWriteMode (you use one handle for both Mdirections of the Socket). Also, it will be initialized with LineBuffering. LUnlike other handlers, the handle will be closed when the function returns. MTherefore, if you are doing threading, you should to it before you call this  handler.  Port Number Server options Handler function Name of logger to use Priority of logged messages Handler to call after logging Resulting handler "Handler to call in the new thread Resulting handler Handler to call  portable provisionalBjrn BringertNone$Parse a date string as formatted by . The resulting & will only have those fields set that I are represented by a format specifier in the format string, and those > fields will be set to the values given in the date string. @ If the same field is specified multiple times, the rightmost  occurence takes precedence. BThe resulting date is not neccessarily a valid date. For example, B if there is no day of the week specifier in the format string,  the value of  will most likely be invalid. >Format specifiers are % followed by some character. All other E characters are treated literally. Whitespace in the format string 9 matches zero or more arbitrary whitespace characters. @Format specifiers marked with * are matched, but do not set any  field in the output. <Some of the format specifiers are marked as space-padded or > zero-padded. Regardless of this, space-padded, zero-padded < or unpadded inputs are accepted. Note that strings using ; unpadded fields without separating the fields may cause  strange parsing. Supported format specfiers: $%% a % character. %a locale')s abbreviated weekday name (Sun ... Sat) %A locale')s full weekday name (Sunday .. Saturday) %b locale'$s abbreviated month name (Jan..Dec) %B locale'&s full month name (January..December) %c locale'6s date and time format (Thu Mar 25 17:47:03 CET 2004) %C century [00-99] %d& day of month, zero padded (01..31) %D date (%m/%d/%y) %e' day of month, space padded ( 1..31) %h same as %b %H- hour, 24-hour clock, zero padded (00..23) %I- hour, 12-hour clock, zero padded (01..12) %j+ day of the year, zero padded (001..366) %k. hour, 24-hour clock, space padded ( 0..23) %l. hour, 12-hour clock, space padded ( 1..12) %m month, zero padded (01..12) %M minute, zero padded (00..59) %n a newline character %p locale's AM or PM indicator %r locale'"s 12-hour time format (hh:mm:ss AM/PM) %R, hours and minutes, 24-hour clock (hh:mm) %s * seconds since '00:00:00 1970-01-01 UTC' %S! seconds, zero padded (00..59) %t a horizontal tab character %T" time, 24-hour clock (hh:mm:ss) %u0 numeric day of the week (1=Monday, 7=Sunday) %U@ * week number, weeks starting on Sunday, zero padded (01-53) %V$ * week number (as per ISO-8601), 7 week 1 is the first week with a Thursday, " zero padded, (01-53) %w1 numeric day of the week, (0=Sunday, 6=Monday) %W@ * week number, weeks starting on Monday, zero padded (01-53) %x locale'%s preferred way of printing dates (%m/%d/%y) %X locale'-s preferred way of printing time. (%H:%M:%S) %y. year, within century, zero padded (00..99) %Y' year, including century. Not padded @ (this is probably a bug, but formatCalendarTime does " it this way). (0-9999) %Z> time zone abbreviation (e.g. CET) or RFC-822 style numeric  timezone (-0500) <Read up to a given number of digits, optionally left-padded , with whitespace and interpret them as an .  Time locale  Date format String to parse  if parsing failed. portable provisional$John Goerzen <jgoerzen@complete.org>None%Split a path in components. Repeated "/" characters don't lead to empty  components. "."K path components are removed. If the path is absolute, the first component will start with "/". ".."% components are left intact. They can' t be simply Kremoved, because the preceding component might be a symlink. In this case, realpath is probably what you need. FThe case that the path is empty, is probably an error. However, it is  treated like "."*, yielding an empty path components list.  Examples:  slice_path "/" = ["/"] 'slice_path "/foo/bar" = ["/foo","bar"] slice_path "..//./" = [".."] slice_path "." = [] See , realpath,  realpath_s. *Form a path from path components. This isn't the inverse of , since  .  normalises the path. See . 4Normalise a path. This is done by reducing repeated /! characters to one, and removing . path components. ..@ path components are left intact, because of possible symlinks.  =  . ESplit a file name in components. This are the base file name and the Lsuffixes, which are separated by dots. If the name starts with a dot, it is Eregarded as part of the base name. The result is a list of file name Kcomponents. The filename may be a path. In this case, everything up to the Hlast path component will be returned as part of the base file name. The path gets normalised thereby. BNo empty suffixes are returned. If the file name contains several Fconsecutive dots, they are regared as part of the preceding file name  component. BConcateneting the name components and adding dots, reproduces the 'original name, with a normalised path: concat . intersperse "." .  ==  normalise. +Note that the last path component might be "..". Then it is not (possible to deduce the refered directory's name from the path. An IO 4action for getting the real path is then necessary.  Examples:    "a.b//./ .foo.tar.gz" == ["a.b/.foo","tar","gz"]  ".x..y." == [".x.", "y."] See , slice_filename'. This is a variant of  . It is like  , except for Lbeing more efficient, and the filename must not contain any preceding path, since this case isn't considered. See , . FForm file name from file name components, interspersing dots. This is the inverse of (, except for normalisation of any path.  - unslice_filename = concat . intersperse "." See . CSplit a path in directory and file name. Only in the case that the Asupplied path is empty, both parts are empty strings. Otherwise, "." is filled in Dfor the corresponding part, if necessary. Unless the path is empty, Iconcatenating the returned path and file name components with a slash in )between, makes a valid path to the file.  split_path* splits off the last path component. This isn'&t the same as the text after the last /. +Note that the last path component might be "..". Then it is not (possible to deduce the refered directory'!s name from the path. Then an IO /action for getting the real path is necessary.  Examples:  *split_path "/a/b/c" == ("/a/b", "c") )split_path "foo" == (".", "foo") +split_path "foo/bar" == ("foo", "bar") *split_path "foo/.." == ("foo", "..") 'split_path "." == (".", ".") %split_path "" == ("", "") )split_path "/foo" == ("/", "foo") )split_path "foo/" == (".", "foo") )split_path "foo/." == (".", "foo") +split_path "foo///./bar" == ("foo", "bar") See . "Get the directory part of a path.  dir_part = fst . split_path See . 'Get the last path component of a path.  !filename_part = snd . split_path  Examples:  !filename_part "foo/bar" == "bar" filename_part "." == "." See .  Inverse of , except for normalisation. /This concatenates two paths, and takes care of "."< and empty paths. When the two components are the result of  split_path, then  unsplit_path Dcreates a normalised path. It is best documented by its definition:  unsplit_path (".", "") = "." unsplit_path ("", ".") = "." unsplit_path (".", q) = q unsplit_path ("", q) = q unsplit_path (p, "") = p unsplit_path (p, ".") = p 'unsplit_path (p, q) = p ++ "/" ++ q  Examples:   unsplit_path ("", "") == "" !unsplit_path (".", "") == "." !unsplit_path (".", ".") == "." #unsplit_path ("foo", ".") == "foo" See . 4Split a file name in prefix and suffix. If there isn't any suffix in Ithe file name, then return an empty suffix. A dot at the beginning or at 1the end is not regarded as introducing a suffix. 8The last path component is what is being split. This isn't the same as Esplitting the string at the last dot. For instance, if the file name doesn'2t contain any dot, dots in previous path component's aren' t mistaken as introducing suffixes. :The path part is returned in normalised form. This means, "." components are removed, and multiple "/"s are reduced to one. Note that there isn'Ht any plausibility check performed on the suffix. If the file name doesn'9t have a suffix, but happens to contain a dot, then this )dot is mistaken as introducing a suffix.  Examples:  Usplit_filename "path/to/foo.bar" = ("path/to/foo","bar") Rsplit_filename "path/to/foo" = ("path/to/foo","") Ssplit_filename "/path.to/foo" = ("/path.to/foo","") Jsplit_filename "a///./x" = ("a/x","") Psplit_filename "dir.suffix/./" = ("dir","suffix") rsplit_filename "Photographie, Das 20. Jahrhundert (300 dpi)" = ("Photographie, Das 20", " Jahrhundert (300 dpi)") See , 'split_filename\''  Variant of #. This is a more efficient version of +, for the case that you know the string is )is a pure file name without any slashes. See .  Inverse of (. Concatenate prefix and suffix, adding Na dot in between, iff the suffix is not empty. The path part of the prefix is  normalised. See . 6Split a path in directory, base file name and suffix. ;Form path from directory, base file name and suffix parts. 4Test a path for a specific suffix and split it off. 5If the path ends with the suffix, then the result is Just prefix, where prefix is the normalised path  without the suffix. Otherwise it's Nothing. ;Make a path absolute, using the current working directory. @This makes a relative path absolute with respect to the current <working directory. An absolute path is returned unmodified. 1The current working directory is determined with getCurrentDirectory Cwhich means that symbolic links in it are expanded and the path is #normalised. This is different from pwd. Make a path absolute. @This makes a relative path absolute with respect to a specified 4directory. An absolute path is returned unmodified. Make a path absolute. @This makes a relative path absolute with respect to a specified 4directory. An absolute path is returned unmodified. CThe order of the arguments can be confusing. You should rather use .  absolute_path'* is included for backwards compatibility.  Guess the ".."-component free form of a path, specified as a list of path components, by syntactically removing them, along with the preceding & path components. This will produce S erroneous results when the path contains symlinks. If the path contains leading ".." components, or more ".."# components than preceeding normal  components, then the ".." components can'2t be normalised away. In this case, the result is Nothing.  Guess the ".."T-component free, normalised form of a path. The transformation is purely syntactic. ".."( path components will be removed, along ; with their preceding path components. This will produce S erroneous results when the path contains symlinks. If the path contains leading ".." components, or more ".."# components than preceeding normal  components, then the ".." components can'2t be normalised away. In this case, the result is Nothing. Cguess_dotdot = fmap unslice_path . guess_dotdot_comps . slice_path %The path to be broken to components. List of path components. List of path components 8The path which consists of the supplied path components Path to be normalised Path in normalised form Path /List of components the file name is made up of File name without path /List of components the file name is made up of List of file name components ;Name of the file which consists of the supplied components Path to be split RDirectory and file name components of the path. The directory path is normalized. Directory and file name 3Path formed from the directory and file name parts )Path including the file name to be split ?The normalised path with the file prefix, and the file suffix. Filename to be split Base name and the last suffix File name prefix and suffix Path Path to split 4Directory part, base file name part and suffix part 4Directory part, base file name part and suffix part (Path consisting of dir, base and suffix Suffix to split off  Path to test Prefix without the suffix or Nothing The path to be made absolute  Absulte path :The directory relative to which the path is made absolute The path to be made absolute Absolute path The path to be made absolute :The directory relative to which the path is made absolute Absolute path List of path components +In case the path could be transformed, the "..")-component free list of path components. Path to be normalised 7In case the path could be transformed, the normalised, "..""-component free form of the path. portable provisional%John Goerzen <jgoerzen@complete.org> None The options for  and  *The base from which calculations are made /The increment to the power for each new suffix -The first power for which suffixes are given The suffixes themselves IPredefined definitions for byte measurement in groups of 1024, from 0 to 2**80 DPredefined definitions for SI measurement, from 10**-24 to 10**24. ATakes a number and returns a new (quantity, suffix) combination. ?The space character is used as the suffix for items around 0. Like 4, but takes a list of numbers. The first number in Kthe list will be evaluated for the suffix. The same suffix and scale will 9be used for the remaining items in the list. Please see  for an example of how this works. 6It is invalid to use this function on an empty list. GRender a number into a string, based on the given quantities. This is Nuseful for displaying quantities in terms of bytes or in SI units. Give this  function the 4 for the desired output, and a precision (number of Hdigits to the right of the decimal point), and you get a string output. Here are some examples:  / Data.Quantity> renderNum binaryOpts 0 1048576  "1M" 0 Data.Quantity> renderNum binaryOpts 2 10485760  "10.00M" / Data.Quantity> renderNum binaryOpts 3 1048576  "1.000M" / Data.Quantity> renderNum binaryOpts 3 1500000  "1.431M" 3 Data.Quantity> renderNum binaryOpts 2 (1500 ** 3)  "3.14G"  ( Data.Quantity> renderNum siOpts 2 1024  "1.02k" + Data.Quantity> renderNum siOpts 2 1048576  "1.05M" ) Data.Quantity> renderNum siOpts 2 0.001  "1.00m" * Data.Quantity> renderNum siOpts 2 0.0001  "100.00u" .If you want more control over the output, see . Like 7, but operates on a list of numbers. The first number Nin the list will be evaluated for the suffix. The same suffix and scale will 2be used for the remaining items in the list. See  for more  examples.  Also, unlike 5, the %f instead of %g printf format is used so that " scientific"$ notation is avoided in the output.  Examples: E *Data.Quantity> renderNums binaryOpts 3 [1500000, 10240, 104857600]  ["1.431M","0.010M","100.000M"] B *Data.Quantity> renderNums binaryOpts 3 [1500, 10240, 104857600] $ ["1.465K","10.000K","102400.000K"] 'Parses a String, possibly generated by . Parses the suffix @and applies it to the number, which is read via the Read class.  Returns Left  error message0 on error, or Right number on successful parse. 9If you want an Integral result, the convenience function  is for you. Parse a number as with , but return the result as an ;. Any type such as Integer, Int, etc. can be used for the  result type. This function simply calls  on the result of . A > is used internally for the parsing of the numeric component. LBy using this function, a user can still say something like 1.5M and get an integral result. Precision of the result The number to examine Prevision of the result The numbers to examine Result &Information on how to parse this data ,Whether to perform a case-insensitive match The string to parse &Information on how to parse this data ,Whether to perform a case-insensitive match The string to parse  portable provisional%John Goerzen <jgoerzen@complete.org> None@January 1, 1970, midnight, UTC, represented as a CalendarTime. SConverts the specified CalendarTime (see System.Time) to seconds-since-epoch time. IThis conversion does respect the timezone specified on the input object. IIf you want a conversion from UTC, specify ctTZ = 0 and ctIsDST = False. HWhen called like that, the behavior is equivolent to the GNU C function (timegm(). Unlike the C library, Haskell's CalendarTime supports Jtimezone information, so if such information is specified, it will impact  the result. :Converts the specified CalendarTime (see System.Time) to seconds-since-epoch format. HThe input CalendarTime is assumed to be the time as given in your local Btimezone. All timezone and DST fields in the object are ignored. KThis behavior is equivolent to the timelocal() and mktime() functions that !C programmers are accustomed to. LPlease note that the behavior for this function during the hour immediately Lbefore or after a DST switchover may produce a result with a different hour than you expect. EConverts the given timeDiff to the number of seconds it represents. 6Uses the same algorithm as normalizeTimeDiff in GHC. JConverts an Epoch time represented with an arbitrary Real to a ClockTime. FThis input could be a CTime from Foreign.C.Types or an EpochTime from System.Posix.Types. FConverts a ClockTime to something represented with an arbitrary Real. NThe result could be treated as a CTime from Foreign.C.Types or EpochTime from $System.Posix.Types. The inverse of . ;Fractions of a second are not preserved by this function. FRender a number of seconds as a human-readable amount. Shows the two (most significant places. For instance:  renderSecs 121 = "2m1s"  See also * for a function that works on a TimeDiff. Like 4, but takes a TimeDiff instead of an integer second count. portable provisional%John Goerzen <jgoerzen@complete.org> NoneLets you examine the  that is contained  within a  object. You can simply pass  a  object and a function to , and   will lock the  object (blocking any E modifications while you are reading it), then pass the object : to your function. If you happen to already have a  L object, withStatus will also accept it and simply pass it unmodified  to the function. The main Progress object. "The main progress status record. An identifying string ;The type for a callback function for the progress tracker. When given at creation time to 'newProgress\'' or when added via , Ithese functions get called every time the status of the tracker changes. This function is passed two  records: the first Areflects the status prior to the update, and the second reflects the status after the update. Please note that the owning ! object will be locked while the Ncallback is running, so the callback will not be able to make changes to it. 8A function that, when called, yields the current time. The default is .  Create a new ' object with the given name and number Iof total units initialized as given. The start time will be initialized Kwith the current time at the present moment according to the system clock. IThe units completed will be set to 0, the time source will be set to the ;system clock, and the parents and callbacks will be empty. If you need more control, see 'newProgress\''.  Example: & prog <- newProgress "mytracker" 1024  Create a new / object initialized with the given status and  callbacks. No adjustment to the  will be made. If you 1want to use the system clock, you can initialize  with the return value of  and also pass  as the timing source. $Adds an new callback to an existing . The callback will be called whenever the object'4s status is updated, except by the call to finishP. KPlease note that the Progress object will be locked while the callback is Krunning, so the callback will not be able to make any modifications to it. !Adds a new parent to an existing . The parent Ewill automatically have its completed and total counters incremented /by the value of those counters in the existing . CCall this when you are finished with the object. It is especially 7important to do this when parent objects are involved. IThis will simply set the totalUnits to the current completedUnits count, Abut will not call the callbacks. It will additionally propogate =any adjustment in totalUnits to the parents, whose callbacks will be called. NThis ensures that the total expected counts on the parent are always correct. JWithout doing this, if, say, a transfer ended earlier than expected, ETA Mvalues on the parent would be off since it would be expecting more data than actually arrived. *Increment the completed unit count in the  object Dby the amount given. If the value as given exceeds the total, then ?the total will also be raised to match this value so that the )completed count never exceeds the total. IYou can decrease the completed unit count by supplying a negative number here. Like , but never modify the total. $Set the completed unit count in the  object to the specified value. Unlike 4, this function sets the count to a specific value, Lrather than adding to the existing value. If this value exceeds the total, Mthen the total will also be raised to match this value so that the completed  count never exceeds teh total. Like , but never modify the total. &Increment the total unit count in the  object by the amount Pgiven. This would rarely be needed, but could be needed in some special cases 9when the total number of units is not known in advance.  Set the total unit count in the  object to the specified  value. Like  , this would rarely be needed. AReturns the speed in units processed per time unit. (If you are Jusing the default time source, this would be units processed per second). 9This obtains the current speed solely from analyzing the  object. 'If no time has elapsed yet, returns 0. "You can use this against either a  object or a  Kobject. This is in the IO monad because the speed is based on the current time.  Example:   getSpeed progressobj >>= print Don'Jt let the type of this function confuse you. It is a fancy way of saying that it can take either a  or a  object, and returns Ka number that is valid as any Fractional type, such as a Double, Float, or  Rational. ?Returns the estimated time remaining, in standard time units. Returns 0 whenever  would return 0. See the comments under $ for information about this function's type  and result. CReturns the estimated system clock time of completion, in standard /time units. Returns the current time whenever  would return 0. See the comments under $ for information about this function's type  and result. =The default time source for the system. This is defined as: -getClockTime >>= (return . clockTimeToEpoch) )Name of this tracker Total units expected The child object  The parent to add to this child   portable provisional%John Goerzen <jgoerzen@complete.org> None,The primary type for bin-packing functions. MThese functions take a list of size of bins. If every bin is the same size,  you can pass repeat binSize) to pass an infinite list of bins if the 6same size. Any surplus bins will simply be ignored.  [size] is the sizes of bins ( [(size, obj)] is the sizes and objects # result is Either error or results ,Potential errors returned as Left values by  functions. Calling  > on this value will produce a nice error message suitable for  display.  Other error 9Bin size1 exceeded by at least the given object and size GRan out of bins; attached value is the list of objects that do not fit JPack objects into bins, preserving order. Objects will be taken from the Oinput list one by one, and added to each bin until the bin is full. Work will Mthen proceed on the next bin. No attempt is made to optimize allocations to Fbins. This is the simplest and most naive bin-packing algorithm, but *may not make very good use of bin space. GPack objects into bins. For each bin, start with the largest objects, Kand keep packing the largest object from the remainder until no object can Gbe found to put in the bin. This is substantially more efficient than #, but requires sorting the input.  -Let us use this as part of the Either monad       portable provisional$John Goerzen <jgoerzen@complete.org>NoneIParse a Comma-Separated Value (CSV) file. The return value is a list of @lines; each line is a list of cells; and each cell is a String. NPlease note that CSV files may have a different number of cells on each line. NAlso, it is impossible to distinguish a CSV line that has a call with no data #from a CSV line that has no cells. Here are some examples:  CInput (literal strings) Parses As (Haskell String syntax) C-------------------------------- --------------------------------- 31,2,3 [["1", "2", "3"]]  2l1 [["l1"], ["l2"]] l2  ( (empty line) [[""]]  4NQ,"Quoted" [["NQ", "Quoted"]]  =NQ,"Embedded""Quote" [["NQ", "Embedded\"Quote"]] "To parse a String, you might use:  %import Text.ParserCombinators.Parsec import Data.String.CSV .... parse csvFile "" mystring (To parse a file, you might instead use:  3do result <- parseFromFile csvFile "/path/to/file" 7Please note that the result of parsing will be of type (Either ParseError [[String]]&). A Left result indicates an error. .For more details, see the Parsec information. ;Generate CSV data for a file. The resulting string can be written out to disk directly. 6portable to platforms with POSIX process\/signal tools provisional$John Goerzen <jgoerzen@complete.org>NoneReturn value from , , , or ?. Contains both a ProcessID and the original command that was $executed. If you prefer not to use  on the result of one Eof these pipe calls, you can use (processID ph), assuming ph is your , as a parameter to . Function that created it Like 6, but returns data in lines instead of just a String. .Shortcut for calling lines on the result from . &Note: this function logs as pipeFrom. Not available on Windows. /Read data from a pipe. Returns a Handle and a . ;When done, you must hClose the handle, and then use either  or getProcessStatus on the ". Zombies will result otherwise.  This function logs as pipeFrom. 'Not available on Windows or with Hugs. 4Read data from a pipe. Returns a lazy string and a . EONLY AFTER the string has been read completely, You must call either  or  on the . Zombies will result otherwise. Not available on Windows. !Write data to a pipe. Returns a  and a new Handle to write to. ;When done, you must hClose the handle, and then use either  or getProcessStatus on the ". Zombies will result otherwise. This function logs as pipeTo. Not available on Windows. ,Write data to a pipe. Returns a ProcessID. You must call either  or  on the ProcessID. Zombies will result otherwise. Not available on Windows. Like a combination of  and  ; returns a 3-tuple of (!, Data From Pipe, Data To Pipe). =When done, you must hClose both handles, and then use either  or getProcessStatus on the ". Zombies will result otherwise. MHint: you will usually need to ForkIO a thread to handle one of the Handles;  otherwise, deadlock can result.  This function logs as pipeBoth. Not available on Windows. Like a combination of  and ; forks an IO thread Ito send data to the piped program, and simultaneously returns its output stream. DThe same note about checking the return status applies here as with . Not available on Windows. Uses  to obtain the exit status Lof the given process ID. If the process terminated normally, does nothing. BOtherwise, raises an exception with an appropriate error message. =This call will block waiting for the given pid to terminate. Not available on Windows. GInvokes the specified command in a subprocess, waiting for the result. EIf the command terminated successfully, return normally. Otherwise, %raises a userError with the problem. Implemented in terms of 8 where supported, and System.Posix.rawSystem otherwise. GInvokes the specified command in a subprocess, waiting for the result. FReturn the result status. Never raises an exception. Only available on POSIX platforms. KLike system(3), this command ignores SIGINT and SIGQUIT and blocks SIGCHLD during its execution. )Logs as System.Cmd.Utils.posixRawSystem CInvokes the specified command in a subprocess, without waiting for Lthe result. Returns the PID of the subprocess -- it is YOUR responsibility Oto use getProcessStatus or getAnyProcessStatus on that at some point. Failure ;to do so will lead to resource leakage (zombie processes). AThis function does nothing with signals. That too is up to you. (Logs as System.Cmd.Utils.forkRawSystem &Open a pipe to the specified command. 0Passes the handle on to the specified function. The 7 specifies what you will be doing. That is, specifing  sets up a pipe from stdin, and  sets up a pipe from stdout. Not available on Windows. -Runs a command, redirecting things to pipes. Not available on Windows. ENote that you may not use the same fd on more than one item. If you 2want to redirect stdout and stderr, dup it first. -Runs a command, redirecting things to pipes. Not available on Windows. 6Returns immediately with the PID of the child. Using  waitProcess on it is YOUR responsibility! ENote that you may not use the same fd on more than one item. If you 2want to redirect stdout and stderr, dup it first. Send stdin to this fd Get stdout from this fd Get stderr from this fd Command to run  Command args Action to run in parent 1Action to run in child before execing (if you don't need something, set this to  return ()) -- IGNORED IN HUGS Send stdin to this fd Get stdout from this fd Get stderr from this fd Command to run  Command args 1Action to run in child before execing (if you don't need something, set this to  return ()) -- IGNORED IN HUGS portable provisional$John Goerzen <jgoerzen@complete.org>None,Transmits an e-mail message using the system's mail transport agent. MThis function takes a message, a list of recipients, and an optional sender, !and transmits it using the system's MTA, sendmail. If sendmail is on the PATH/, it will be used; otherwise, a list of system $default locations will be searched. 3A failure will be logged, since this function uses   internally. This function will first try sendmail%. If it does not exist, an error is  logged under System.Cmd.Utils.pOpen3 and various default sendmail locations Mare tried. If that still fails, an error is logged and an exception raised. >The envelope from address. If not specified, takes the system's default, which is usually based on the effective userid of the current process. This is not necessarily what you want, so I recommend specifying it. CA list of recipients for your message. An empty list is an error. The message itself. portable provisional%John Goerzen <jgoerzen@complete.org> NoneFTakes a IO action and a function. The IO action will be called in a Nseparate thread. When it is completed, the specified function is called with 7its result. This is a simple way of doing callbacks. portable provisional%John Goerzen <jgoerzen@complete.org> None$Main class for readable mailboxes. The mailbox object a represents zero or more  s. Each message has a unique identifier b- in a format specific to each given mailbox. .This identifier may or may not be persistent. IFunctions which return a list are encouraged -- but not guaranteed -- to do so lazily. /Implementing classes must provide, at minimum,  . +Returns a list of all unique identifiers.  @Returns a list of all unique identifiers as well as all flags.  9Returns a list of all messages, including their content, ' flags, and unique identifiers.  .Returns information about specific messages.  .A Message is represented as a simple String.  Convenience shortcut /The flags which may be assigned to a message.                     portable provisional%John Goerzen <jgoerzen@complete.org> NoneGenerate (return) a . Retrieve the next token from a  stream. A The given function should return the value to use, or Nothing  to cause an error. A shortcut to 0; the test here is just a function that returns Ja Bool. If the result is true; return that value -- otherwise, an error. ,Matches one item in a list and returns it. $Matches all items and returns them 0Matches one item not in a list and returns it. ,Matches one specific token and returns it. Running notMatching p msg will try to apply parser p. CIf it fails, returns (). If it succeds, cause a failure and raise Ethe given error message. It will not consume input in either case.    6portable to platforms with POSIX process\/signal tools provisional$John Goerzen <jgoerzen@complete.org>None ADetach the process from a controlling terminal and run it in the =background, handling it with standard Unix deamon semantics. <After running this, please note the following side-effects: , The PID of the running process will change . stdin, stdout, and stderr will not work (they' ll be set to  /dev/null)  CWD will be changed to / I highly< suggest running this function before starting any threads. CNote that this is not intended for a daemon invoked from inetd(1).      portable provisional$John Goerzen <jgoerzen@complete.org>None'!0The writing side of a Haskell pipe. Please see G for more  details. "0The reading side of a Haskell pipe. Please see G for more  details. #A # simulates true I/(O, but uses an in-memory buffer instead of on-disk storage.  It provides ,a full interface like Handle (it implements  HVIOReader,  HVIOWriter, and  HVIOSeeker7). However, it maintains an in-memory buffer with the Jcontents of the file, rather than an actual on-disk file. You can access Jthe entire contents of this buffer at any time. This can be quite useful  for testing I/.O code, or for cases where existing APIs use I/ O, but you 2prefer a String representation. You can create a # with a call to D.  The present #4 implementation is rather inefficient, particularly 0when reading towards the end of large files. It's best used for smallish 6data storage. This problem will be fixed eventually. $ Simulate I/O based on a string buffer. When a $8 is created, it is initialized based on the contents of a  C. Its contents are read lazily whenever a request is made to read something from the $. It Lcan be used, therefore, to implement filters (simply initialize it with the Mresult from, say, a map over hGetContents from another HVIO object), codecs,  and simple I/<O testing. Because it is lazy, it need not hold the entire $string in memory. You can create a $ with a call to C. %This is the generic I/2O support class. All objects that are to be used /in the HVIO system must provide an instance of %. LFunctions in this class provide an interface with the same specification as Hthe similar functions in System.IO. Please refer to that documentation 9for a more complete specification than is provided here.  Instances of % must provide &, * , and either ' or (. 9Implementators of readable objects must provide at least 0 and 4. An implementation of 2! is also highly suggested, since ?the default cannot implement proper partial closing semantics. 9Implementators of writable objects must provide at least 5 and :. 9Implementators of seekable objects must provide at least >, <, and ;. & Close a file 'Test if a file is open (Test if a file is closed )(Raise an error if the file is not open. < This is a new HVIO function and is implemented in terms of  '. *Whether or not we'(re at EOF. This may raise on exception @ on some items, most notably write-only Handles such as stdout. @ In general, this is most reliable on items opened for reading. 8 vIsEOF implementations must implicitly call vTestOpen. +Detailed show output. ,Make an IOError. -Throw an IOError. .Get the filename/object/#whatever that this corresponds to.  May be Nothing. /Throw an isEOFError if we'&re at EOF; returns nothing otherwise. ? If an implementation overrides the default, make sure that it ? calls vTestOpen at some point. The default implementation is  a wrapper around a call to *. 0Read one character 1Read one line 2@Get the remaining contents. Please note that as a user of this G function, the same partial-closing semantics as are used in the  standard ! are  encouraged from implementators,  but are not required". That means that, for instance,  a 0 after a 2 may return some undefined G result instead of the error you would normally get. You should / use caution to make sure your code doesn't fall into that trap, @ or make sure to test your code with Handle or one of the M default instances defined in this module. Also, some implementations @ may essentially provide a complete close after a call to 2. ( The bottom line: after a call to 2, you should do nothing 1 else with the object save closing it with &. CFor implementators, you are highly encouraged to provide a correct  implementation. 39Indicate whether at least one item is ready for reading. < This will always be True for a great many implementations. 4=Indicate whether a particular item is available for reading. 5Write one character 6Write a string 7/Write a string with newline character after it 8?Write a string representation of the argument, plus a newline. 9Flush any output buffers. D Note: implementations should assure that a vFlush is automatically  performed A on file close, if necessary to ensure all data sent is written. :AIndicate whether or not this particular object supports writing. ;Seek to a specific location. <Get the current position. =@Convenience function to reset the file pointer to the beginning  of the file. A call to  vRewind h is the  same as ; h AbsoluteSeek 0. >1Indicate whether this instance supports seeking. ?.Set buffering; the default action is a no-op. @>Get buffering; the default action always returns NoBuffering. AGBinary output: write the specified number of octets from the specified  buffer location. B;Binary input: read the specified number of octets from the / specified buffer location, continuing to read @ until it either consumes that much data or EOF is encountered. C Returns the number of octets actually read. EOF errors are never 9 raised; fewer bytes than requested are returned on EOF. C Create a new $ object. D Create a new #& instance. The buffer is initialized Mto the value passed, and the pointer is placed at the beginning of the file. -You can put things in it by using the normal 6 calls, and reset to "the beginning by using the normal = call. The function is called when &* is called, and is passed the contents of 'the buffer at close time. You can use E if you don' t want to  do anything. 2To create an empty buffer, pass the initial value "". E,Default (no-op) memory buf close function. F4Grab the entire contents of the buffer as a string. Unlike 2/, this has no effect on the open status of the Ditem, the EOF status, or the current position of the file pointer. GCreate a Haskell pipe. &These pipes are analogous to the Unix 3pipes that are available from System.Posix, but don't require Unix and work Monly in Haskell. When you create a pipe, you actually get two HVIO objects: a " and a !. You must use the ! in one thread and the " in another thread. Data that's written to the !- will then be available for reading with the ". The Mpipes are implemented completely with existing Haskell threading primitives, Kand require no special operating system support. Unlike Unix pipes, these Kpipes cannot be used across a fork(). Also unlike Unix pipes, these pipes 6are portable and interact well with Haskell threads. B!""#$%&#'$()%&'()*+,-./0123456789:;<=>?@AB*+,-CInitial contents of the $ .DInitial Contents  close func E/FG0123456789:;<'!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFG'%&'()*+,-./0123456789:;<=>?@AB$C#DEF"!G!""#$&%#'$()%&'()*+,-./0123456789:;<=>?@AB*+,-C.DE/FG0123456789:;<portable provisional%John Goerzen <jgoerzen@complete.org> None JETypes that can open a HVIO object should be instances of this class. You need only implement K. OThe main HVFS class. 9Default implementations of these functions are provided:  ] -- implemented in terms of [  ^  S -- implemented in terms of [  T -- implemented in terms of [  U -- implemented in terms of \  \ -- set to call [. /Default implementations of all other functions Hwill generate an isIllegalOperation error, since they are assumed to be un-implemented. %You should always provide at least a [ call, and almost !certainly several of the others. GMost of these functions correspond to functions in System.Directory or FSystem.Posix.Files. Please see detailed documentation on them there. U8True if the file exists, regardless of what type it is. A This is even True if the given path is a broken symlink. ^3Raise an error relating to actions on this class. b6Evaluating types of files and information about them. HThis corresponds to the System.Posix.Types.FileStatus type, and indeed, $that is one instance of this class. +Inplementators must, at minimum, implement r and q. CDefault implementations of everything else are provided, returning reasonable values. FA default implementation of this is not currently present on Windows. e@Refers to file permissions, NOT the st_mode field from stat(2) u Similar to w , but for K result. wEncapsulate a b* result. This is required due to Haskell .typing restrictions. You can get at it with:  case encap of $ HVFSStatEncap x -> -- now use x yBConvenience function for working with stat -- takes a stat result 6and a function that uses it, and returns the result. )Here is an example from the HVFS source:  " vGetModificationTime fs fp = $ do s <- vGetFileStatus fs fp C return $ epochToClockTime (withStat s vModificationTime) See  for more information. z Similar to y, but for the K result. =Error handler helper 8HIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz=>?@A= HIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz=OPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstJKLMNuvwxyzHI  HIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz=>?@A!,portable to platforms supporting binary I\/O provisional$John Goerzen <jgoerzen@complete.org>None{<Provides support for handling binary blocks with convenient types. ?This module provides implementations for Strings and for [Word8] (lists of  Word8s). ~*As a wrapper around the standard function B, 'this function takes a standard Haskell   instead of the far less  convenient Ptr a5. The entire contents of the string will be written as a binary buffer using B$. The length of the output will be )the length of the passed String or list. =If it helps, you can thing of this function as being of type Handle -> String -> IO ()  An alias for ~ C ,Acts a wrapper around the standard function D, :this function returns a standard Haskell String (or [Word8] ) instead of  modifying a 'Ptr a'; buffer. The length is the maximum length to read and the semantice are the same as with D; namely, the empty string Cis returned with EOF is reached, and any given read may read fewer bytes than the given length.  (Actually, it's a wrapper around B)  An alias for  E Like 5, but guarantees that it will only return fewer than 8the requested number of bytes when EOF is encountered.  An alias for  E FGWrites the list of blocks to the given file handle -- a wrapper around ~. Think of this function as:  Handle -> [String] -> IO () (You can use it that way)  An alias for F C )putBlocks :: (BinaryConvertible b) => [[b]] -> IO () putBlocks = hPutBlocks stdout AReturns a lazily-evaluated list of all blocks in the input file,  as read by 1. There will be no 0-length block in this list. The list simply ends at EOF.  An alias for  E Same as  , but using  underneath.  An alias for  E CBinary block-based interaction. This is useful for scenarios that Etake binary blocks, manipulate them in some way, and then write them out. Take a look at ' for an example. The integer argument 8is the size of input binary blocks. This function uses   internally.  An alias for  over E and C Same as  , but uses  instead of  internally.  An alias for  over E and C JCopies everything from the input handle to the output handle using binary 7blocks of the given size. This was once the following beautiful implementation:  8 hBlockCopy bs hin hout = hBlockInteract bs hin hout id (GF is the built-in Haskell function that just returns whatever is given to it) KIn more recent versions of MissingH, it uses a more optimized routine that 8avoids ever having to convert the binary buffer at all.  Copies from E to C( using binary blocks of the given size.  An alias for  over E and C /Copies one filename to another in binary mode. DPlease note that the Unix permission bits on the output file cannot *be set due to a limitation of the Haskell H Gfunction. Therefore, you may need to adjust those bits after the copy  yourself. #This function is implemented using  internally. Like the built-in I', but opens the file in binary instead of text mode. JSame as  , but works with HVFS objects. Like the built-in K', but opens the file in binary instead of text mode. LLike , but works on HVFS objects. {|}~FMNJLOP{|}~{|}~{|}~FMNJLOP"portable provisional%John Goerzen <jgoerzen@complete.org> None'Obtain a recursive listing of all files/directories beneath 7the specified directory. The traversal is depth-first and the original -item is always present in the returned list. 9If the passed value is not a directory, the return value be only that value. The "." and ".."- entries are removed from the data returned. Like 8, but return the stat() (System.Posix.Files.FileStatus) Ninformation with them. This is an optimization if you will be statting files yourself later. The items are returned lazily. NWARNING: do not change your current working directory until you have consumed 6all the items. Doing so could cause strange effects. GAlternatively, you may wish to pass an absolute path to this function. ERemoves a file or a directory. If a directory, also removes all its  child files/ directories. @Provide a result similar to the command ls -l over a directory. GKnown bug: setuid bit semantics are inexact compared with standard ls. HIHI#portable provisional$John Goerzen <jgoerzen@complete.org>None DGiven a list of strings, output a line containing each item, adding Mnewlines as appropriate. The list is not expected to have newlines already. @Given a handle, returns a list of all the lines in that handle. KThanks to lazy evaluation, this list does not have to be read all at once. Combined with *, this can make a powerful way to develop filters. See the $ function for more on that concept.  Example:  main = do  l <- hGetLines stdin 6 hPutStrLns stdout $ filter (startswith "1") l  This is similar to the built-in Q , but works *on any handle, not just stdin and stdout. In other words: # interact = hInteract stdin stdout :Line-based interaction. This is similar to wrapping your interact functions with R and S. This equality holds:  + lineInteract = hLineInteract stdin stdout Here's an example:  / main = lineInteract (filter (startswith "1")) IThis will act as a simple version of grep -- all lines that start with 1 /will be displayed; all others will be ignored. @Line-based interaction over arbitrary handles. This is similar to wrapping hInteract with R and S. (One could view this function like this:  % hLineInteract finput foutput func = * let newf = unlines . func . lines in ' hInteract finput foutput newf 9Though the actual implementation is this for efficiency: % hLineInteract finput foutput func =  do  lines <- hGetLines finput % hPutStrLns foutput (func lines) 5Copies from one handle to another in raw mode (using hGetContents). DCopies from one handle to another in raw mode (using hGetContents). :Takes a function to provide progress updates to the user. =Copies from one handle to another in text mode (with lines). Like  hBlockCopy, this implementation is nice: 0 hLineCopy hin hout = hLineInteract hin hout id  Copies from E to C using lines. An alias for  over E and C. -Copies one filename to another in text mode. HPlease note that the Unix permission bits are set at a default; you may -need to adjust them after the copy yourself. #This function is implemented using  internally. ISets stdin and stdout to be block-buffered. This can save a huge amount Mof system resources since far fewer syscalls are made, and can make programs run much faster. ASets stdin and stdout to be line-buffered. This saves resources Kon stdout, but not many on stdin, since it it still looking for newlines. >Applies a given function to every item in a list, and returns  the new list. Unlike the system'%s mapM, items are evaluated lazily.  Input handle Output handle LProgress function -- the bool is always False unless this is the final call    $portable provisional%John Goerzen <jgoerzen@complete.org> None  The type used for functions for . See  for details. 4Merge two sorted lists into a single, sorted whole.  Example:  , merge [1,3,5] [1,2,4,6] -> [1,1,2,3,4,5,6] QuickCheck test property: prop_merge xs ys = 1 merge (sort xs) (sort ys) == sort (xs ++ ys) " where types = xs :: [Int] :Merge two sorted lists using into a single, sorted whole, <allowing the programmer to specify the comparison function. QuickCheck test property: prop_mergeBy xs ys = I mergeBy cmp (sortBy cmp xs) (sortBy cmp ys) == sortBy cmp (xs ++ ys) + where types = xs :: [ (Int, Int) ] 2 cmp (x1,_) (x2,_) = compare x1 x2 CReturns true if the given list starts with the specified elements; (false otherwise. (This is an alias for Data.List.isPrefixOf.)  Example: ! startswith "He" "Hello" -> True AReturns true if the given list ends with the specified elements; (false otherwise. (This is an alias for Data.List.isSuffixOf.)  Example:  endswith "lo" "Hello" -> True JReturns true if the given list contains any of the elements in the search list. GSimilar to Data.List.takeWhile, takes elements while the func is true. =The function is given the remainder of the list to examine. GSimilar to Data.List.dropWhile, drops elements while the func is true. =The function is given the remainder of the list to examine. ISimilar to Data.List.span, but performs the test on the entire remaining #list instead of just one element.  spanList p xs is the same as ((takeWhileList p xs, dropWhileList p xs) JSimilar to Data.List.break, but performs the test on the entire remaining "list instead of just one element. AGiven a delimiter and a list (or string), split into components.  Example:  < split "," "foo,bar,,baz," -> ["foo", "bar", "", "baz", ""] 5 split "ba" ",foo,bar,,baz," -> [",foo,","r,,","z,"] KGiven a list and a replacement list, replaces each occurance of the search 6list with the replacement list in the operation list.  Example:  +replace "," "." "127,0,0,1" -> "127.0.0.1" 'This could logically be thought of as: -replace old new l = join new . split old $ l CGiven a delimiter and a list of items (or strings), join the items by using the delimiter.  Example: 1 join "|" ["foo", "bar", "baz"] -> "foo|bar|baz" Like 9, but works with a list of anything showable, converting it to a String.  Examples: / genericJoin ", " [1, 2, 3, 4] -> "1, 2, 3, 4" D genericJoin "|" ["foo", "bar", "baz"] -> "\"foo\"|\"bar\"|\"baz\"" DReturns true if the given parameter is a sublist of the given list; false otherwise.  Example:  5 contains "Haskell" "I really like Haskell." -> True / contains "Haskell" "OCaml is great." -> False 6This function was submitted to GHC and was applied as T2. This function therefore is deprecated and will be removed in future versions. EAdds the specified (key, value) pair to the given list, removing any 2existing pair with the same key already present. ARemoves all (key, value) pairs from the given list where the key matches the given one. GReturns the keys that comprise the (key, value) pairs of the given AL.  Same as: map fst DReturns the values the comprise the (key, value) pairs of the given AL.  Same as: map snd 6Indicates whether or not the given key is in the AL. DFlips an association list. Converts (key1, val), (key2, val) pairs to (val, [key1, key2]). @Converts an association list to a string. The string will have Pone pair per line, with the key and value both represented as a Haskell string. 8This function is designed to work with [(String, String)] association lists, (but may work with other types as well. The inverse of /, this function reads a string and outputs the appropriate association list. Like 1, this is designed to work with [(String, String)] association Hlists but may also work with other objects with simple representations. HReturns a count of the number of times the given element occured in the  given list. 8Returns the rightmost index of the given element in the  given list. 4Like elemRIndex, but returns -1 if there is nothing found. +Forces the evaluation of the entire list. BThis is an enhanced version of the concatMap or map functions in  Data.List. "Unlike those functions, this one: E Can consume a varying number of elements from the input list during  each iteration 5 Can arbitrarily decide when to stop processing data H Can return a varying number of elements to insert into the output list 5 Can actually switch processing functions mid-stream < Is not even restricted to processing the input list intact 'The function used by wholeMap, of type , is repeatedly called Fwith the input list. The function returns three things: the function Ito call for the next iteration (if any), what remains of the input list, Mand the list of output elements generated during this iteration. The return  value of 7 is the concatenation of the output element lists from all iterations. EProcessing stops when the remaining input list is empty. An example of a  is . DA parser designed to process fixed-width input fields. Use it with . MThe Int list passed to this function is the list of the field widths desired Lfrom the input. The result is a list of those widths, if possible. If any Mof the input remains after processing this list, it is added on as the final Pelement in the result list. If the input is less than the sum of the requested Owidths, then the result list will be short the appropriate number of elements, 5and its final element may be shorter than requested.  Examples: -wholeMap (fixedWidth [1, 2, 3]) "1234567890"  --> ["1","23","456","7890"] -wholeMap (fixedWidth (repeat 2)) "123456789"  --> ["12","34","56","78","9"] %wholeMap (fixedWidth []) "123456789"  --> ["123456789"] <wholeMap (fixedWidth [5, 3, 6, 1]) "Hello, This is a test." , --> ["Hello",", T","his is"," ","a test."] 7Helps you pick out fixed-width components from a list.  Example:  "conv :: String -> (String,String) conv = runState $  do f3 <- grab 3  n2 <- grab 2 $ return $ f3 ++ "," ++ n2  main = print $ conv "TestIng" Prints: ("Tes,tI","ng") ISimilar to Data.List.elemIndex. Instead of looking for one element in a Llist, this function looks for the first occurance of a sublist in the list, Nand returns the index of the first element of that occurance. If there is no such list, returns Nothing. IIf the list to look for is the empty list, will return Just 0 regardless &of the content of the list to search.  Examples: %subIndex "foo" "asdfoobar" -> Just 3 subIndex "foo" [] -> Nothing subIndex "" [] -> Just 0 subIndex "" "asdf" -> Just 0 ,subIndex "test" "asdftestbartest" -> Just 4 7subIndex [(1::Int), 2] [0, 5, 3, 2, 1, 2, 4] -> Just 4 FGiven a list, returns a new list with all duplicate elements removed.  For example:  uniq "Mississippi" -> "Misp" JYou should not rely on this function necessarily preserving order, though 'the current implementation happens to. 5This function is not compatible with infinite lists. -This is presently an alias for Data.List.nub !List of elements to look for List to search Result !! %portable provisional%John Goerzen <jgoerzen@complete.org> None<Converts a String, String Map into a string representation. See & for more on the similar function for 3association lists. This implementation is simple:  'strFromM = strFromAL . Data.Map.toList BThis function is designed to work with Map String String objects, Cbut may also work with other objects with simple representations. 2Converts a String into a String, String Map. See 2 for more on the similar function for association lists. This implementation is simple:  %strToM = Data.Map.fromList . strToAL BThis function is designed to work with Map String String objects, but may work with other key/'value combinations if they have simple representations. Flips a Map. See  for more on the similar function for lists. >Returns a list of all keys in the Map whose value matches the =parameter. If the value does not occur in the Map, the empty list is returned. BPerforms a lookup, and raises an exception (with an error message @prepended with the given string) if the key could not be found. &portable provisional$John Goerzen <jgoerzen@complete.org>NoneESplits a pathname into a tuple representing the root of the name and Othe extension. The extension is considered to be all characters from the last Ldot after the last slash to the end. Either returned string may be empty. HMake an absolute, normalized version of a path with all double slashes, !dot, and dotdot entries removed. LThe first parameter is the base for the absolut calculation; in many cases, 6it would correspond to the current working directory. OThe second parameter is the pathname to transform. If it is already absolute,  the first parameter is ignored.  Nothing may be returned if there'#s an error; for instance, too many .. entries for the given path. ELike absNormPath, but returns Nothing if the generated result is not 1the passed base path or a subdirectory thereof. ,Creates a temporary directory for your use. OThe passed string should be a template suitable for mkstemp; that is, end with "XXXXXX". ?Your string should probably start with the value returned from (System.Directory.getTemporaryDirectory. 4The name of the directory created will be returned. /Creates a temporary directory for your use via , @runs the specified action (passing in the directory name), then Premoves the directory and all its contents when the action completes (or raises an exception. 9Changes the current working directory to the given path, executes the given I/7O action, then changes back to the original directory,  even if the I/O action raised an exception. Runs the given I/0O action with the CWD set to the given tmp dir, Eremoving the tmp dir and changing CWD back afterwards, even if there was an exception. .Absolute path for use with starting directory The path name to make absolute Result .Absolute path for use with starting directory The path to make absolute   'portable provisional$John Goerzen <jgoerzen@complete.org>None An in-memory read/0write filesystem. Think of it as a dynamically 'resizable ramdisk written in Haskell. (The content of a file or directory in a . The basic node of a +. The String corresponds to the filename,  and the entry to the contents.  A simple System.IO.HVFS.HVFSStat 4class that assumes that everything is either a file or a directory. !True if file, False if directory #Set to 0 if unknown or a directory  Create a new  object from an existing tree. - An empty filesystem may be created by using [] for the parameter.  Create a new  object using an IORef to an  existing tree.  Similar to &0 but the first element won't be /. nice_slice "/" -> [] (nice_slice "/foo/bar" -> ["foo", "bar"] /Gets a full path, after investigating the cwd. Gets the full path via , then splits it via . U8Find an element on the tree, assuming a normalized path V8Find an element on the tree, normalizing the path first WXYUVZ[\]WXYUVZ[\](portable provisional$John Goerzen <jgoerzen@complete.org>NoneAAccess a subdirectory of a real filesystem as if it was the root of that filesystem. <Restrict access to the underlying filesystem to be strictly ;read-only. Any write-type operations will cause an error. %No constructor is required; just say HVFSReadOnly fs to make a !new read-only wrapper around the O instance fs.  Create a new  object. ^Get the embedded object _/Convert a local (chroot) path to a full path. `/Convert a full path to a local (chroot) path. /Convert a local (chroot) path to a full path. abc"The object to pass requests on to 'The path of the directory to make root The resulting new object ^_`defgh abc^_`defgh)portable provisional$John Goerzen <jgoerzen@complete.org>None !Return value from guessing a file's type. IThe first element of the tuple gives the MIME type. It is Nothing if no suitable type could be found. QThe second element gives the encoding. It is Nothing if there was no particular 9encoding for the file, or if no encoding could be found. :A mapping used to expand common suffixes into equivolent, ( better-parsed versions. For instance, .tgz would expand  into .tar.gz. 4A mapping used to determine the encoding of a file. $ This is used, for instance, to map .gz to gzip. 0A mapping used to map extensions to MIME types. A mapping used to augment the  when non-strict  lookups are used. ARead the given mime.types file and add it to an existing object. Returns new object. 5Load a mime.types file from an already-open handle. <Guess the type of a file given a filename or URL. The file 0 is not opened; only the name is considered. 6Guess the extension of a file based on its MIME type. . The return value includes the leading dot. 0Returns Nothing if no extension could be found. >In the event that multiple possible extensions are available, E one of them will be picked and returned. The logic to select one - of these should be considered undefined.  Similar to ., but returns a list of all possible matching 8extensions, or the empty list if there are no matches. i@Adds a new type to the data structures, replacing whatever data K may exist about it already. That is, it overrides existing information K about the given extension, but the same type may occur more than once. Default MIME type data to use Read the system'7s default mime.types files, and add the data contained 8therein to the passed object, then return the new one. Data to work with Whether to work on strict data  File to read  New object Data to work with Whether to work on strict data Handle to read from  New object Source data for guessing  Whether to limit to strict data File or URL name to consider Result of guessing (see ! for details on interpreting it) Source data for guessing  Whether to limit to strict data MIME type to consider 4Result of guessing, or Nothing if no match possible Source data for guessing  Whether to limit to strict data MIME type to consider Result of guessing i Source data "Whether to add to strict data set MIME type to add Extension to add Result of addition jklmnopijklmnop*portable provisional$John Goerzen <jgoerzen@complete.org>None@Removes any whitespace characters that are present at the start >or end of a string. Does not alter the internal contents of a Dstring. If no whitespace characters are present at the start or end Dof a string, returns the original string unmodified. Safe to use on  any string. 2Note that this may differ from some other similar &functions from other authors in that:  If multiple whitespace ;characters are present all in a row, they are all removed;  If no 4whitespace characters are present, nothing is done. Same as 3, but applies only to the left side of the string. Same as 4, but applies only to the right side of the string. ASplits a string around whitespace. Empty elements in the result !list are automatically removed. FEscape all characters in the input pattern that are not alphanumeric. 4Does not make special allowances for NULL, which isn' t valid in a %Haskell regular expression pattern. 8Attempts to parse a value from the front of the string. q  q+portable provisional$John Goerzen <jgoerzen@complete.org>Noner,The main data type for the progress meter. s The master  object for overall status tIndividual component statuses uWidth of the meter vUnits of display wFunction to render numbers xAuto-updating display (Set up a new status bar using defaults:  The given tracker  Width 80 ' Data.Quantity.renderNums binaryOpts 1  Unit inticator B Set up a new status bar. &Adjust the list of components of this . 0Add a new component to the list of components. Remove a component by name. Adjusts the width of this . GLike renderMeter, but prints it to the screen instead of returning it. .This function will output CR, then the meter. >Pass stdout as the handle for regular display to the screen. ?Clears the meter -- outputs CR, spaces equal to the width - 1, then another CR. >Pass stdout as the handle for regular display to the screen. DClears the meter, writes the given string, then restores the meter. 5The string is assumed to contain a trailing newline. >Pass stdout as the handle for regular display to the screen. BStarts a thread that updates the meter every n seconds by calling the specified function. Note: displayMeter stdout is an ideal function here. ,Save this threadID and use it later to call stopAutoDisplayMeter. +Stops the specified meter from displaying. You should probably call  after a call to this. Render the current status. rystuvwxThe top-level  Unit indicator string $Width of the terminal -- usually 80 A function to render sizes zThe meter to display Update interval in seconds Function to display it Resulting thread id {  rystuvwxz{,portable provisional$John Goerzen <jgoerzen@complete.org>None:Convert a wildcard to an (uncompiled) regular expression. FCheck the given name against the given pattern, being case-sensitive. OThe given pattern is forced to match the given name starting at the beginning. (The wildcard pattern to use as the base !The filename to check against it Result |}|}-portable provisional$John Goerzen <jgoerzen@complete.org>NoneCTakes a pattern. Returns a list of names that match that pattern. The pattern is evaluated by System.Path.WildMatch. This function :does not perform tilde or environment variable expansion. JFilenames that begin with a dot are not included in the result set unless 6that component of the pattern also begins with a dot. *In MissingH, this function is defined as: glob = vGlob SystemFS Like  , but works on both the system (real) and HVFS virtual filesystems. ~~.portable provisional$John Goerzen <jgoerzen@complete.org>None"Main parser for the control file Dependency parser. 1Returns (package name, Maybe version, arch list)  version is (operator, operand)  12312412512612712812912:1;<1=>?@ABCDEFGHIJKLMNOPPQRSTUVWXYZ[\]^_`abcdefghijk l m n o p p q q r r s s k t u v w x y z { y | } ~     |        1      !"#$%&'()*+,-./0123456789:;<=>?@ABCDE F G H I J K L M N O P Q R S T U V W X Y Z [ \ ] ^ _ ` a b c d e f g h i j k lmmnopqrstuvwxyz{|}~!!!!!!!!!!!!!!!!!!!!!!""""#############$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$%%%%%&&&&&&&'''''''''X'''''(((()))))))))))) ) ) * * ****+++++++++++++,, -!-".#.$%&'D(12)12)12*12*12+12+12,12,12-12-12.12.12/12/12012012112112212212312312412412512512612612712812912:12;<=><=?<=@<=A<=B<=C<=D<=E<=F<=G<=H<=I<=J<=K<=L<=M<=N<=O<PQ<PR<PS<PT<Pc<Pb<Pa<P`<P_<P^<P]<PU<PV<PW<P[<PZ<PY<PX<PW<PV<PU<PT<PS<PR<PQ<PX<PY<PN<PO<PL<PK<PJ<PI<PH<PM<PZ<P[<P\<P]<P^<P_<P`<Pa<Pb<Pc<Pd<Pe<Pf<Pg<Ph<Pi<Pj<Pk<Pl<Pm<Pn<Po<Pp<Pq<Pr<Ps<Pt<Pu<Pv<Pw<Px<Py<Pz<P{<P|}~e                                %111%1   <  11 F G    H I  ! " # $ % & ' ( ) * + , - . / 0 1 2 345678191:;1<1:=!>1?1:@1AB!C1AD!E!F!G!H!I1AJ1KL1KM1KN'O'P''Q'R'S'T'U'V(W(X(Y((Z([(\(](^(_(`)a)b)c)d)e)f)g)h*i+j+k+l+m+n+o+p+j+q+r,s,t-u-v-w..x.y.z.{.|.}~MissingH-1.2.0.2System.IO.HVFSSystem.Console.GetOpt.Utils System.DebianSystem.Posix.ConstsSystem.IO.StatCompatSystem.IO.PlafCompatData.Compression.InflateData.Hash.MD5.Zord64_HARD Data.Hash.MD5Data.Hash.CRC32.GZipData.Hash.CRC32.PosixData.Bits.UtilsSystem.FileArchive.GZipData.Tuple.UtilsData.Maybe.UtilsData.Either.Utils Network.UtilsNetwork.SocketServerSystem.Time.ParseDateSystem.Path.NameManip Data.QuantitySystem.Time.UtilsData.Progress.TrackerData.BinPackingData.CSVSystem.Cmd.UtilsNetwork.Email.SendmailControl.Concurrent.Thread.UtilsNetwork.Email.Mailbox#Text.ParserCombinators.Parsec.Utils System.DaemonSystem.IO.HVIOSystem.IO.BinarySystem.IO.HVFS.UtilsSystem.IO.UtilsData.List.UtilsData.Map.Utils System.PathSystem.IO.HVFS.InstanceHelpersSystem.IO.HVFS.CombinatorsData.MIME.TypesData.String.UtilsData.Progress.MeterSystem.Path.WildMatchSystem.Path.GlobSystem.Debian.ControlParserSystem.IO.WindowsCompat NameManipbaseSystem.Posix.Types LinkCountUserIDGroupID EpochTimeDeviceIDFileIDFileMode FileOffset GHC.IO.IOModeIOModeGHC.IOFilePath StdOption parseCmdLinevalidateCmdLine stdRequired stdOptional DebVersion ControlFilecompareDebVersioncheckDebVersionblockSpecialModecharacterSpecialMode namedPipeModeregularFileMode directoryMode fileTypeModes socketModesymbolicLinkModeFileStatusCompatdeviceIDfileIDfileMode linkCount fileOwner fileGroupspecialDeviceIDfileSize accessTimemodificationTimestatusChangeTime sc_helper isBlockDeviceisCharacterDevice isNamedPipe isRegularFile isDirectoryisSymbolicLinkisSocket nullFileNameBitOutputinflate_stringinflate_string_remainderbits_to_word32inflateZord64MD5get_nextlen_padfinishedWordListBoolListStrABCDmd5md5smd5i update_crcupdate_crc_list calc_crc32 gzipcrctab iter_crc32crc32crctabgetBytes fromBytesc2w8s2w8w82cw82sSectionFootersize crc32validHeadermethodflagsextrafilenamecommentmtimexflos GZipError UnknownError UnknownMethod NotGZIPFileCRCError hDecompress decompress read_sections read_section read_headerfst3snd3thd3 forceMaybe forceMaybeMsg maybeToEither forceEitherforceEitherMsgeitherToMonadErrorfromLeft fromRight fromEither niceSocketsDo connectTCPconnectTCPAddr listenTCPAddr showSockAddr SocketServer optionsSSsockSSHandlerTInetServerOptionslistenQueueSize portNumber interfacereusefamilysockTypeprotoStrsimpleTCPOptionssetupSocketServercloseSocketServer handleOne serveForeverserveTCPforeverloggingHandlerthreadedHandler handleHandlerparseCalendarTime slice_path unslice_pathnormalise_pathslice_filenameslice_filename'unslice_filename split_pathdir_part filename_part unsplit_pathsplit_filenamesplit_filename'unsplit_filenamesplit3unsplit3 test_suffix absolute_pathabsolute_path_byabsolute_path'guess_dotdot_comps guess_dotdotSizeOpts powerIncr firstPowersuffixes binaryOptssiOpts quantifyNum quantifyNums renderNum renderNumsparseNum parseNumIntepochtimegm timelocaltimeDiffToSecsepochToClockTimeclockTimeToEpoch renderSecsrenderTDProgressStatuses withStatusProgressProgressStatuscompletedUnits totalUnits startTime trackerName timeSourceProgressCallbackProgressTimeSource newProgress newProgress' addCallback addParentfinishPincrPincrP'setPsetP' incrTotalsetTotalgetSpeedgetETRgetETAdefaultTimeSource BinPackerBinPackerErrorBPOtherBPSizeTooLarge BPTooFewBins packByOrderpackLargeFirstcsvFile genCsvFile PipeHandle processID phCommandphArgs phCreatorPipeMode WriteToPipe ReadFromPipe pipeLinesFrom hPipeFrompipeFromhPipeTopipeTo hPipeBothpipeBoth forceSuccess safeSystemposixRawSystem forkRawSystempOpenpOpen3 pOpen3Rawsendmail runInThread MailboxWriterappendMessagesdeleteMessagesaddFlags removeFlagssetFlags MailboxReaderlistIDslistMessageFlagsgetAll getMessagesMessageFlagsFlag OTHERFLAG FORWARDEDDRAFTDELETEDFLAGGEDANSWEREDSEENGeneralizedTokenParserGeneralizedTokentogtoktokengsatisfygoneOfgallgnoneOfg specificg notMatching detachDaemon PipeWriter PipeReader MemoryBuffer StreamReaderHVIOvClosevIsOpen vIsClosed vTestOpenvIsEOFvShow vMkIOErrorvThrowvGetFPvTestEOFvGetCharvGetLine vGetContentsvReady vIsReadablevPutCharvPutStr vPutStrLnvPrintvFlush vIsWritablevSeekvTellvRewind vIsSeekable vSetBuffering vGetBufferingvPutBufvGetBufnewStreamReadernewMemoryBuffermbDefaultCloseFuncgetMemoryBuffer newHVIOPipeSystemFS HVFSOpenablevOpen vReadFile vWriteFilevOpenBinaryFileHVFSvGetCurrentDirectoryvSetCurrentDirectoryvGetDirectoryContentsvDoesFileExistvDoesDirectoryExist vDoesExistvCreateDirectoryvRemoveDirectoryvRenameDirectory vRemoveFile vRenameFilevGetFileStatusvGetSymbolicLinkStatusvGetModificationTime vRaiseErrorvCreateSymbolicLinkvReadSymbolicLink vCreateLinkHVFSStat vDeviceIDvFileID vFileMode vLinkCount vFileOwner vFileGroupvSpecialDeviceID vFileSize vAccessTimevModificationTimevStatusChangeTimevIsBlockDevicevIsCharacterDevice vIsNamedPipevIsRegularFile vIsDirectoryvIsSymbolicLink vIsSocket HVFSOpenEncap HVFSStatEncapwithStatwithOpenBinaryConvertibletoBuffromBuf hPutBufStr putBufStr hGetBufStr getBufStrhFullGetBufStr fullGetBufStr hGetBlocks getBlockshFullGetBlocks fullGetBlockshBlockInteract blockInteracthFullBlockInteractfullBlockInteract hBlockCopy blockCopycopyFileBlocksToFilereadBinaryFilewriteBinaryFile recurseDirrecurseDirStatrecursiveRemovelsl hPutStrLns hGetLines hInteract lineInteract hLineInteracthCopy hCopyProgress hLineCopylineCopycopyFileLinesToFileoptimizeForBatchoptimizeForInteractionlazyMapM WholeFuncmergemergeBy startswithendswithhasAny takeWhileList dropWhileListspanList breakListsplitreplacejoin genericJoincontainsaddToAL delFromALkeysALvaluesALhasKeyALflipAL strFromALstrToAL countElem elemRIndexalwaysElemRIndexseqListwholeMap fixedWidthgrabsubIndexuniqstrFromMstrToMflipMflippedLookupM forceLookupMsplitExt absNormPathsecureAbsNormPathmktmpdir brackettmpdir bracketCWDbrackettmpdirCWD MemoryVFS MemoryEntry MemoryFileMemoryDirectory MemoryNode SimpleStatisFile newMemoryVFSnewMemoryVFSRef nice_slice getFullPath getFullSlice HVFSChroot HVFSReadOnly newHVFSChroot MIMEResults MIMETypeData suffixMap encodingsMaptypesMapcommonTypesMap readMIMETypeshReadMIMETypes guessTypeguessExtensionguessAllExtensions defaultmtdreadSystemMIMETypesstriplstriprstripsplitWsescapeRe maybeRead ProgressMetersimpleNewMeternewMeter setComponents addComponentremoveComponentsetWidth displayMeter clearMeterwriteMeterStringautoDisplayMeterkillAutoDisplayMeter renderMeter wildToRegex wildCheckCaseglobvGlobcontroldepPartghc-prim GHC.ClassesOrd$fOrdDebVersionCDevCInoCModeCOffCPidCSsizeCGidCNlinkCUidCCcCSpeedCTcflagCRLimFd ByteCount ClockTick ProcessIDProcessGroupIDLimit unix-2.6.0.1System.Posix.Files getPathVar setFileSize touchFile setFileTimessetSymbolicLinkOwnerAndGroupsetOwnerAndGrouprenamereadSymbolicLinkcreateSymbolicLink removeLink createLink createDevicecreateNamedPipegetSymbolicLinkStatus getFileStatus fileExist fileAccess setFileModeSystem.Posix.Files.Common getFdPathVar setFdSizesetFdOwnerAndGroup getFdStatusstatusChangeTimeHiResmodificationTimeHiResaccessTimeHiRessetFileCreationMask setFdModeintersectFileModesunionFileModes accessModes otherModes groupModes ownerModes stdFileModesetGroupIDMode setUserIDModeotherExecuteModeotherWriteMode otherReadModegroupExecuteModegroupWriteMode groupReadModeownerExecuteModeownerWriteMode ownerReadMode nullFileMode FileStatus FileSizeBits LinkLimitInputLineLimitInputQueueLimit FileNameLimit PathNameLimitPipeBufferLimitSymbolicLinkLimitSetOwnerAndGroupIsRestrictedFileNamesAreNotTruncated VDisableCharAsyncIOAvailablePrioIOAvailableSyncIOAvailablePathVarTreeNullLeafBranchInfMStatebitsoffsethistorylocTablesTableLengthLitLenDistCodeshow_b int_to_bits word8_to_bitsset_bits align_8_bitsget_bits extract_InfM output_w32 repeat_w32s get_word32sget_w32get_bitinflate_blocksinflate_tablestriplemake_lit_dist_lengths meta_code inflate_codeslitlens dist_codedistsmk_basesinflate_trees_fixed make_tableget_code make_tree $fMonadInfM $fShowBitW64lohi w64ToInteger integerToW64 $fEnumZord64 $fRealZord64$fIntegralZord64 $fBitsZord64 $fNumZord64 $fReadZord64 $fShowZord64RotationXYZmd5_main md5_do_block md5_roundmd5_inner_functionmd5_fmd5_gmd5_hmd5_i magic_numbersroundsabcd_to_stringdisplay_32bits_as_hexabcd_to_integerrev_numstring_to_word32sbools_to_word32slength_to_chars $fNumABCD $fMD5WordList$fMD5Str $fMD5BoolListmagicfFHCRCfFEXTRAfFNAME fFCOMMENT read_datasplit1 parseword$fErrorGZipErrorold-time-1.1.0.1 System.TimeformatCalendarTime CalendarTimectWDayreadN GHC.TypesInt Data.MaybeNothing pCalendarTime parseEnumsetYearsetMonthsetDaysetHoursetMinsetSecsetWDaysetYDay setTZNamesetTZ updateYear updateHour setHour12read2read3intsignGHC.RealIntegralroundDoubleProgressRecords withRecordProgressRecordparents callbacksstatusnow modStatus callParents!$fProgressStatusesProgressStatusb$fProgressRecordsProgressIO$fProgressStatusesProgressIOGHC.Showshow$fErrorBinPackerErrorpackLargeFirst'$fShowBinPackerErroreolcell quotedchar quotedcelllineSystem.Posix.Process.CommongetProcessStatuslogbase logRunningwarnFail cmdfailed cmdsignalledshowCmd sendmails findsendmailsendmail_workertrapchild1child2GHC.BaseStringGHC.IO.Handle.Text hGetContentsPipeBitPipeEOFVIOCloseSupport vioc_isopenvioc_get vioc_closevioc_setsrvvrvprvpr_getcpwvpwmv$fHVIOPipeWriter$fShowPipeWriter$fHVIOPipeReader$fShowPipeReader$fHVIOMemoryBuffer$fShowMemoryBuffer$fHVIOStreamReader$fShowStreamReader $fHVIOHandleeh$fHVFSOpenableSystemFS$fHVFSSystemFS$fHVFSStatFileStatus$fShowFileStatushPutBufGHC.IO.Handle.FDstdouthGetBufstdin hPutBlocksidopenBinaryFile System.IOreadFilevReadBinaryFile writeFilevWriteBinaryFilehGetBlocksUtilhBlockInteractUtil$fBinaryConvertibleWord8$fBinaryConvertibleCharinteract Data.Listlinesunlines isInfixOf findMelemgetMelemcontentcwd$fHVFSOpenableMemoryVFS$fHVFSMemoryVFS$fShowMemoryVFS$fHVFSStatSimpleStatdchdch2fpfp2dchwithroroerrordch2fph$fHVFSOpenableHVFSChroot$fHVFSHVFSChroot$fHVFSOpenableHVFSReadOnly$fHVFSHVFSReadOnlyaddType getStrict setStrictdefaultfilelocationsdefault_encodings_mapdefault_suffix_mapdefault_types_mapdefault_common_typeswscharsProgressMeterRmasterP componentswidthunitrendererautoDisplayers clearmeterstr renderMeterRconvwildconvpathasWild expandGlobrunGlobextlineentry headerPGP blankLine headerHashheader