darcs-2.14.1: a distributed, interactive, smart revision control system

Copyright (c) The University of Glasgow 2001David Roundy 2003-2005 GPL (I'm happy to also license this file BSD style but don't want to bother distributing two license files with darcs. droundy@abridgegame.org experimental portable None Haskell2010

Darcs.Util.ByteString

Description

GZIp and MMap IO for ByteStrings, encoding utilities, and miscellaneous functions for Data.ByteString

Synopsis

# IO with mmap or gzip

Read an entire file, which may or may not be gzip compressed, directly into a ByteString.

Like readFilePS, this reads an entire file directly into a ByteString, but it is even more efficient. It involves directly mapping the file to memory. This has the advantage that the contents of the file never need to be copied. Also, under memory pressure the page may simply be discarded, wile in the case of readFilePS it would need to be written to swap. If you read many small files, mmapFilePS will be less memory-efficient than readFilePS, since each mmapFilePS takes up a separate page of memory. Also, you can run into bus errors if the file is modified.

Read standard input, which may or may not be gzip compressed, directly into a ByteString.

type FileSegment = (FilePath, Maybe (Int64, Int)) Source #

Pointer to a filesystem, possibly with start/end offsets. Supposed to be fed to (uncurry mmapFileByteString) or similar.

Read in a FileSegment into a Lazy ByteString. Implemented using mmap.

# gzip handling

Decompress the given bytestring into a lazy list of chunks, along with a boolean flag indicating (if True) that the CRC was corrupted. Inspecting the flag will cause the entire list of chunks to be evaluated (but if you throw away the list immediately this should run in constant space).

# list utilities

Drop leading white space, where white space is defined as consisting of ' ', '\t', '\n', or '\r'.

Split at first occurrence of ' ', '\t', '\n', or '\r'.

readIntPS skips any whitespace at the beginning of its argument, and reads an Int from the beginning of the PackedString. If there is no integer at the beginning of the string, it returns Nothing, otherwise it just returns the int read, along with a B.ByteString containing the remainder of its input.

Return the B.ByteString between the two lines given, or Nothing if they do not appear.

O(n) The intercalate function takes a ByteString and a list of ByteStrings and concatenates the list after interspersing the first argument between each element of the list.

# encoding and unicode utilities

Test if a ByteString is made of ascii characters

Decode a ByteString to a String according to the current locale, using lone surrogates for un-decodable bytes.

Encode a String to a ByteString according to the current locale, converting lone surrogates back to the original byte. If that fails (because the locale does not support the full unicode range) then encode using utf-8, assuming that the un-ecodable characters come from patch meta data.

See also setEnvCautiously.

Decode a ByteString containing UTF-8 to a String. Decoding errors are flagged with the U+FFFD character.

Encode a String to a ByteString using UTF-8.

# properties

Simpler but less efficient variant of betweenLinesPS.