This manual is for Lzlib (version 1.11, 2 January 2019).
Copyright © 2009-2019 Antonio Diaz Diaz.
This manual is free documentation: you have unlimited permission to copy, distribute and modify it.
Lzlib is a data compression library providing in-memory LZMA compression and decompression functions, including integrity checking of the decompressed data. The compressed data format used by the library is the lzip format. Lzlib is written in C.
The lzip file format is designed for data sharing and long-term archiving, taking into account both data integrity and decoder availability:
A nice feature of the lzip format is that a corrupt byte is easier to repair the nearer it is from the beginning of the file. Therefore, with the help of lziprecover, losing an entire archive just because of a corrupt byte near the beginning is a thing of the past.
The functions and variables forming the interface of the compression library are declared in the file 'lzlib.h'. Usage examples of the library are given in the files 'main.c' and 'bbexample.c' from the source distribution.
Compression/decompression is done by repeatedly calling a couple of read/write functions until all the data have been processed by the library. This interface is safer and less error prone than the traditional zlib interface.
Compression/decompression is done when the read function is called. This means the value returned by the position functions will not be updated until a read call, even if a lot of data are written. If you want the data to be compressed in advance, just call the read function with a size equal to 0.
If all the data to be compressed are written in advance, lzlib will automatically adjust the header of the compressed data to use the largest dictionary size that does not exceed neither the data size nor the limit given to 'LZ_compress_open'. This feature reduces the amount of memory needed for decompression and allows minilzip to produce identical compressed output as lzip.
Lzlib will correctly decompress a data stream which is the concatenation of two or more compressed data streams. The result is the concatenation of the corresponding decompressed data streams. Integrity testing of concatenated compressed data streams is also supported.
Lzlib is able to compress and decompress streams of unlimited size by automatically creating multimember output. The members so created are large, about 2 PiB each.
All the library functions are thread safe. The library does not install any signal handler. The decoder checks the consistency of the compressed data, so the library should never crash even in case of corrupted input.
In spite of its name (Lempel-Ziv-Markov chain-Algorithm), LZMA is not a concrete algorithm; it is more like "any algorithm using the LZMA coding scheme". For example, the option '-0' of lzip uses the scheme in almost the simplest way possible; issuing the longest match it can find, or a literal byte if it can't find a match. Inversely, a much more elaborated way of finding coding sequences of minimum size than the one currently used by lzip could be developed, and the resulting sequence could also be coded using the LZMA coding scheme.
Lzlib currently implements two variants of the LZMA algorithm; fast (used by option '-0' of minilzip) and normal (used by all other compression levels).
The high compression of LZMA comes from combining two basic, well-proven compression ideas: sliding dictionaries (LZ77/78) and markov models (the thing used by every compression algorithm that uses a range encoder or similar order-0 entropy coder as its last stage) with segregation of contexts according to what the bits are used for.
The ideas embodied in lzlib are due to (at least) the following people: Abraham Lempel and Jacob Ziv (for the LZ algorithm), Andrey Markov (for the definition of Markov chains), G.N.N. Martin (for the definition of range encoding), Igor Pavlov (for putting all the above together in LZMA), and Julian Seward (for bzip2's CLI).
LANGUAGE NOTE: Uncompressed = not compressed = plain data; it may never have been compressed. Decompressed is used to refer to data which have undergone the process of decompression.
The application should compare LZ_version and LZ_version_string for consistency. If the first character differs, the library code actually used may be incompatible with the 'lzlib.h' header file used by the application.
if( LZ_version() != LZ_version_string ) error( "bad library version" );
Lzlib internal functions need access to a memory chunk at least as large as the dictionary size (sliding window). For efficiency reasons, the input buffer for compression is twice or sixteen times as large as the dictionary size.
Finally, for safety reasons, lzlib uses two more internal buffers.
These are the four buffers used by lzlib, and their guaranteed minimum sizes:
These functions provide minimum and maximum values for some parameters. Current values are shown in square brackets.
Returns the base 2 logarithm of the smallest valid dictionary size .
Returns the smallest valid dictionary size [4 KiB].
Returns the base 2 logarithm of the largest valid dictionary size .
Returns the largest valid dictionary size [512 MiB].
Returns the largest valid match length limit .
These are the functions used to compress data. In case of error, all of them return -1 or 0, for signed and unsigned return values respectively, except 'LZ_compress_open' whose return value must be verified by calling 'LZ_compress_errno' before using it.
Initializes the internal stream state for compression and returns a pointer that can only be used as the encoder argument for the other LZ_compress functions, or a null pointer if the encoder could not be allocated.
The returned pointer must be verified by calling 'LZ_compress_errno' before using it. If 'LZ_compress_errno' does not return 'LZ_ok', the returned pointer must not be used and should be freed with 'LZ_compress_close' to avoid memory leaks.
dictionary_size sets the dictionary size to be used, in bytes. Valid values range from 4 KiB to 512 MiB. Note that dictionary sizes are quantized. If the specified size does not match one of the valid sizes, it will be rounded upwards by adding up to (dictionary_size / 8) to it.
match_len_limit sets the match length limit in bytes. Valid values range from 5 to 273. Larger values usually give better compression ratios but longer compression times.
If dictionary_size is 65535 and match_len_limit is 16, the fast variant of LZMA is chosen, which produces identical compressed output as
lzip -0. (The dictionary size used will be rounded upwards to 64 KiB).
member_size sets the member size limit in bytes. Valid values range from 100 kB to 2 PiB. Small member size may degrade compression ratio, so use it only when needed. To produce a single-member data stream, give member_size a value larger than the amount of data to be produced. Values larger than 2 PiB will be reduced to 2 PiB to prevent the uncompressed size of the member from overflowing.
Frees all dynamically allocated data structures for this stream. This function discards any unprocessed input and does not flush any pending output. After a call to 'LZ_compress_close', encoder can no longer be used as an argument to any LZ_compress function.
Use this function to tell 'lzlib' that all the data for this member have already been written (with the 'LZ_compress_write' function). It is safe to call 'LZ_compress_finish' as many times as needed. After all the produced compressed data have been read with 'LZ_compress_read' and 'LZ_compress_member_finished' returns 1, a new member can be started with 'LZ_compress_restart_member'.
Use this function to start a new member in a multimember data stream. Call this function only after 'LZ_compress_member_finished' indicates that the current member has been fully read (with the 'LZ_compress_read' function).
Use this function to make available to 'LZ_compress_read' all the data already written with the 'LZ_compress_write' function. First call 'LZ_compress_sync_flush'. Then call 'LZ_compress_read' until it returns 0.
Repeated use of 'LZ_compress_sync_flush' may degrade compression ratio, so use it only when needed.
The 'LZ_compress_read' function reads up to size bytes from the stream pointed to by encoder, storing the results in buffer.
The return value is the number of bytes actually read. This might be less than size; for example, if there aren't that many bytes left in the stream or if more bytes have to be yet written with the 'LZ_compress_write' function. Note that reading less than size bytes is not an error.
The 'LZ_compress_write' function writes up to size bytes from buffer to the stream pointed to by encoder.
The return value is the number of bytes actually written. This might be less than size. Note that writing less than size bytes is not an error.
The 'LZ_compress_write_size' function returns the maximum number of bytes that can be immediately written through the 'LZ_compress_write' function.
It is guaranteed that an immediate call to 'LZ_compress_write' will accept a size up to the returned number of bytes.
Returns the current error code for encoder. See Error codes.
Returns 1 if all the data have been read and 'LZ_compress_close' can be safely called. Otherwise it returns 0. 'LZ_compress_finished' implies 'LZ_compress_member_finished'.
Returns 1 if the current member, in a multimember data stream, has been fully read and 'LZ_compress_restart_member' can be safely called. Otherwise it returns 0.
Returns the number of input bytes already compressed in the current member.
Returns the number of compressed bytes already produced, but perhaps not yet read, in the current member.
Returns the total number of input bytes already compressed.
Returns the total number of compressed bytes already produced, but perhaps not yet read.
These are the functions used to decompress data. In case of error, all of them return -1 or 0, for signed and unsigned return values respectively, except 'LZ_decompress_open' whose return value must be verified by calling 'LZ_decompress_errno' before using it.
Initializes the internal stream state for decompression and returns a pointer that can only be used as the decoder argument for the other LZ_decompress functions, or a null pointer if the decoder could not be allocated.
The returned pointer must be verified by calling 'LZ_decompress_errno' before using it. If 'LZ_decompress_errno' does not return 'LZ_ok', the returned pointer must not be used and should be freed with 'LZ_decompress_close' to avoid memory leaks.
Frees all dynamically allocated data structures for this stream. This function discards any unprocessed input and does not flush any pending output. After a call to 'LZ_decompress_close', decoder can no longer be used as an argument to any LZ_decompress function.
Use this function to tell 'lzlib' that all the data for this stream have already been written (with the 'LZ_decompress_write' function). It is safe to call 'LZ_decompress_finish' as many times as needed.
Resets the internal state of decoder as it was just after opening it with the 'LZ_decompress_open' function. Data stored in the internal buffers is discarded. Position counters are set to 0.
Resets the error state of decoder and enters a search state that lasts until a new member header (or the end of the stream) is found. After a successful call to 'LZ_decompress_sync_to_member', data written with 'LZ_decompress_write' will be consumed and 'LZ_decompress_read' will return 0 until a header is found.
This function is useful to discard any data preceding the first member, or to discard the rest of the current member, for example in case of a data error. If the decoder is already at the beginning of a member, this function does nothing.
The 'LZ_decompress_read' function reads up to size bytes from the stream pointed to by decoder, storing the results in buffer.
The return value is the number of bytes actually read. This might be less than size; for example, if there aren't that many bytes left in the stream or if more bytes have to be yet written with the 'LZ_decompress_write' function. Note that reading less than size bytes is not an error.
In case of decompression error caused by corrupt or truncated data, 'LZ_decompress_read' does not signal the error immediately to the application, but waits until all decoded bytes have been read. This allows tools like tarlz to recover as much data as possible from each damaged member.
The 'LZ_decompress_write' function writes up to size bytes from buffer to the stream pointed to by decoder.
The return value is the number of bytes actually written. This might be less than size. Note that writing less than size bytes is not an error.
The 'LZ_decompress_write_size' function returns the maximum number of bytes that can be immediately written through the 'LZ_decompress_write' function.
It is guaranteed that an immediate call to 'LZ_decompress_write' will accept a size up to the returned number of bytes.
Returns the current error code for decoder. See Error codes.
Returns 1 if all the data have been read and 'LZ_decompress_close' can be safely called. Otherwise it returns 0.
Returns 1 if the previous call to 'LZ_decompress_read' finished reading the current member, indicating that final values for member are available through 'LZ_decompress_data_crc', 'LZ_decompress_data_position', and 'LZ_decompress_member_position'. Otherwise it returns 0.
Returns the version of current member from member header.
Returns the dictionary size of current member from member header.
Returns the 32 bit Cyclic Redundancy Check of the data decompressed from the current member. The returned value is valid only when 'LZ_decompress_member_finished' returns 1.
Returns the number of decompressed bytes already produced, but perhaps not yet read, in the current member.
Returns the number of input bytes already decompressed in the current member.
Returns the total number of input bytes already decompressed.
Returns the total number of decompressed bytes already produced, but perhaps not yet read.
Most library functions return -1 to indicate that they have failed. But this return value only tells you that an error has occurred. To find out what kind of error it was, you need to verify the error code by calling 'LZ_(de)compress_errno'.
Library functions don't change the value returned by 'LZ_(de)compress_errno' when they succeed; thus, the value returned by 'LZ_(de)compress_errno' after a successful call is not necessarily LZ_ok, and you should not use 'LZ_(de)compress_errno' to determine whether a call failed. If the call failed, then you can examine 'LZ_(de)compress_errno'.
The error codes are defined in the header file 'lzlib.h'.
The value of this constant is 0 and is used to indicate that there is no error.
At least one of the arguments passed to the library function was invalid.
No memory available. The system cannot allocate more virtual memory because its capacity is full.
A library function was called in the wrong order. For example 'LZ_compress_restart_member' was called before 'LZ_compress_member_finished' indicates that the current member is finished.
An invalid member header (one with the wrong magic bytes) was read. If this happens at the end of the data stream it may indicate trailing data.
The end of the data stream was reached in the middle of a member.
The data stream is corrupt. If 'LZ_decompress_member_position' is 6 or less, it indicates either a format version not supported, an invalid dictionary size, a corrupt header in a multimember data stream, or trailing data too similar to a valid lzip header. Lziprecover can be used to remove conflicting trailing data from a file.
A bug was detected in the library. Please, report it. See Problems.
Returns the standard error message for a given error code. The messages are fairly short; there are no multi-line messages or embedded newlines. This function makes it easy for your program to report informative error messages about the failure of a library call.
The value of lz_errno normally comes from a call to 'LZ_(de)compress_errno'.
Minilzip is a test program for the lzlib compression library, fully compatible with lzip 1.4 or newer.
The format for running minilzip is:
minilzip [options] [files]
'-' used as a file argument means standard input. It can be mixed with other files and is read just once, the first time it appears in the command line.
minilzip supports the following options:
For maximum compression you should use a dictionary size limit as large
as possible, but keep in mind that the decompression memory requirement
is affected at compression time by the choice of dictionary size limit.
-0 .. -9
The bidimensional parameter space of LZMA can't be mapped to a linear scale optimal for all files. If your files are large, very repetitive, etc, you may need to use the '--dictionary-size' and '--match-length' options directly to achieve optimal performance.
If several compression levels or '-s' or '-m' options are given, the last setting is used. For example '-9 -s64MiB' is equivalent to '-s64MiB -m273'
|Level||Dictionary size (-s)||Match length limit (-m)
|-0||64 KiB||16 bytes
|-1||1 MiB||5 bytes
|-2||1.5 MiB||6 bytes
|-3||2 MiB||8 bytes
|-4||3 MiB||12 bytes
|-5||4 MiB||20 bytes
|-6||8 MiB||36 bytes
|-7||16 MiB||68 bytes
|-8||24 MiB||132 bytes
|-9||32 MiB||273 bytes
Numbers given as arguments to options may be followed by a multiplier and an optional 'B' for "byte".
Table of SI and binary prefixes (unit multipliers):
|k||kilobyte (10^3 = 1000)|||||Ki||kibibyte (2^10 = 1024)
|M||megabyte (10^6)|||||Mi||mebibyte (2^20)
|G||gigabyte (10^9)|||||Gi||gibibyte (2^30)
|T||terabyte (10^12)|||||Ti||tebibyte (2^40)
|P||petabyte (10^15)|||||Pi||pebibyte (2^50)
|E||exabyte (10^18)|||||Ei||exbibyte (2^60)
|Z||zettabyte (10^21)|||||Zi||zebibyte (2^70)
|Y||yottabyte (10^24)|||||Yi||yobibyte (2^80)
Exit status: 0 for a normal exit, 1 for environmental problems (file not found, invalid flags, I/O errors, etc), 2 to indicate a corrupt or invalid input file, 3 for an internal consistency error (eg, bug) which caused minilzip to panic.
Perfection is reached, not when there is no longer anything to add, but
when there is no longer anything to take away.
-- Antoine de Saint-Exupery
In the diagram below, a box like this:
+---+ | | <-- the vertical bars might be missing +---+
represents one byte; a box like this:
+==============+ | | +==============+
represents a variable number of bytes.
A lzip data stream consists of a series of "members" (compressed data sets). The members simply appear one after another in the data stream, with no additional information before, between, or after them.
Each member has the following structure:
+--+--+--+--+----+----+=============+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | ID string | VN | DS | LZMA stream | CRC32 | Data size | Member size | +--+--+--+--+----+----+=============+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
All multibyte values are stored in little endian order.
This chapter shows the order in which the library functions should be called depending on what kind of data stream you want to compress or decompress. See the file 'bbexample.c' in the source distribution for an example of how buffer-to-buffer compression/decompression can be implemented using lzlib.
Note that lzlib's interface is symmetrical. That is, the code for normal compression and decompression is identical except because one calls LZ_compress* functions while the other calls LZ_decompress* functions.
Example 1: Normal compression (member_size > total output).
1) LZ_compress_open 2) LZ_compress_write 3) if no more data to write, call LZ_compress_finish 4) LZ_compress_read 5) go back to step 2 until LZ_compress_finished returns 1 6) LZ_compress_close
Example 2: Normal compression using LZ_compress_write_size.
1) LZ_compress_open 2) go to step 5 if LZ_compress_write_size returns 0 3) LZ_compress_write 4) if no more data to write, call LZ_compress_finish 5) LZ_compress_read 6) go back to step 2 until LZ_compress_finished returns 1 7) LZ_compress_close
Example 3: Decompression.
1) LZ_decompress_open 2) LZ_decompress_write 3) if no more data to write, call LZ_decompress_finish 4) LZ_decompress_read 5) go back to step 2 until LZ_decompress_finished returns 1 6) LZ_decompress_close
Example 4: Decompression using LZ_decompress_write_size.
1) LZ_decompress_open 2) go to step 5 if LZ_decompress_write_size returns 0 3) LZ_decompress_write 4) if no more data to write, call LZ_decompress_finish 5) LZ_decompress_read 5a) optionally, if LZ_decompress_member_finished returns 1, read final values for member with LZ_decompress_data_crc, etc. 6) go back to step 2 until LZ_decompress_finished returns 1 7) LZ_decompress_close
Example 5: Multimember compression (member_size < total output).
1) LZ_compress_open 2) go to step 5 if LZ_compress_write_size returns 0 3) LZ_compress_write 4) if no more data to write, call LZ_compress_finish 5) LZ_compress_read 6) go back to step 2 until LZ_compress_member_finished returns 1 7) go to step 10 if LZ_compress_finished() returns 1 8) LZ_compress_restart_member 9) go back to step 2 10) LZ_compress_close
Example 6: Multimember compression (user-restarted members).
1) LZ_compress_open (with member_size > largest member). 2) LZ_compress_write 3) LZ_compress_read 4) go back to step 2 until member termination is desired 5) LZ_compress_finish 6) LZ_compress_read 7) go back to step 6 until LZ_compress_member_finished returns 1 9) go to step 12 if all input data have been written and LZ_compress_finished returns 1 10) LZ_compress_restart_member 11) go back to step 2 12) LZ_compress_close
Example 7: Decompression with automatic removal of leading data.
1) LZ_decompress_open 2) LZ_decompress_sync_to_member 3) go to step 6 if LZ_decompress_write_size returns 0 4) LZ_decompress_write 5) if no more data to write, call LZ_decompress_finish 6) LZ_decompress_read 7) go back to step 3 until LZ_decompress_finished returns 1 8) LZ_decompress_close
Example 8: Streamed decompression with automatic resynchronization to next member in case of data error.
1) LZ_decompress_open 2) go to step 5 if LZ_decompress_write_size returns 0 3) LZ_decompress_write 4) if no more data to write, call LZ_decompress_finish 5) if LZ_decompress_read produces LZ_header_error or LZ_data_error, call LZ_decompress_sync_to_member 6) go back to step 2 until LZ_decompress_finished returns 1 7) LZ_decompress_close
There are probably bugs in lzlib. There are certainly errors and omissions in this manual. If you report them, they will get fixed. If you don't, no one will ever know about them and they will remain unfixed for all eternity, if not longer.
If you find a bug in lzlib, please send electronic mail to
firstname.lastname@example.org. Include the version number, which you can
find by running
minilzip --version or in
'LZ_version_string' from 'lzlib.h'.