summaryrefslogtreecommitdiffstats
path: root/doc/plzip.texi
blob: e13515aacd9876aea54cbb8bbdbf5ea1434cc6eb (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
\input texinfo @c -*-texinfo-*-
@c %**start of header
@setfilename plzip.info
@documentencoding ISO-8859-15
@settitle Plzip Manual
@finalout
@c %**end of header

@set UPDATED 9 July 2015
@set VERSION 1.4

@dircategory Data Compression
@direntry
* Plzip: (plzip).               Parallel compressor compatible with lzip
@end direntry


@ifnothtml
@titlepage
@title Plzip
@subtitle Parallel compressor compatible with lzip
@subtitle for Plzip version @value{VERSION}, @value{UPDATED}
@author by Antonio Diaz Diaz

@page
@vskip 0pt plus 1filll
@end titlepage

@contents
@end ifnothtml

@node Top
@top

This manual is for Plzip (version @value{VERSION}, @value{UPDATED}).

@menu
* Introduction::           Purpose and features of plzip
* Invoking plzip::         Command line interface
* Program design::         Internal structure of plzip
* File format::            Detailed format of the compressed file
* Memory requirements::    Memory required to compress and decompress
* Minimum file sizes::     Minimum file sizes required for full speed
* Problems::               Reporting bugs
* Concept index::          Index of concepts
@end menu

@sp 1
Copyright @copyright{} 2009-2015 Antonio Diaz Diaz.

This manual is free documentation: you have unlimited permission
to copy, distribute and modify it.


@node Introduction
@chapter Introduction
@cindex introduction

Plzip is a massively parallel (multi-threaded) lossless data compressor
based on the lzlib compression library, with a user interface similar to
the one of lzip, bzip2 or gzip.

Plzip can compress/decompress large files on multiprocessor machines
much faster than lzip, at the cost of a slightly reduced compression
ratio (0.4 to 2 percent larger compressed files). Note that the number
of usable threads is limited by file size; on files larger than a few GB
plzip can use hundreds of processors, but on files of only a few MB
plzip is no faster than lzip (@pxref{Minimum file sizes}).

Plzip uses the lzip file format; the files produced by plzip are fully
compatible with lzip-1.4 or newer, and can be rescued with lziprecover.

The lzip file format is designed for data sharing and long-term
archiving, taking into account both data integrity and decoder
availability:

@itemize @bullet
@item
The lzip format provides very safe integrity checking and some data
recovery means. The
@uref{http://www.nongnu.org/lzip/manual/lziprecover_manual.html#Data-safety,,lziprecover}
program can repair bit-flip errors (one of the most common forms of data
corruption) in lzip files, and provides data recovery capabilities,
including error-checked merging of damaged copies of a file.
@ifnothtml
@ref{Data safety,,,lziprecover}.
@end ifnothtml

@item
The lzip format is as simple as possible (but not simpler). The lzip
manual provides the code of a simple decompressor along with a detailed
explanation of how it works, so that with the only help of the lzip
manual it would be possible for a digital archaeologist to extract the
data from a lzip file long after quantum computers eventually render
LZMA obsolete.

@item
Additionally the lzip reference implementation is copylefted, which
guarantees that it will remain free forever.
@end itemize

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.

Plzip uses the same well-defined exit status values used by lzip and
bzip2, which makes it safer than compressors returning ambiguous warning
values (like gzip) when it is used as a back end for other programs like
tar or zutils.

Plzip will automatically use the smallest possible dictionary size for
each file without exceeding the given limit. Keep in mind that the
decompression memory requirement is affected at compression time by the
choice of dictionary size limit (@pxref{Memory requirements}).

When compressing, plzip replaces every file given in the command line
with a compressed version of itself, with the name "original_name.lz".
When decompressing, plzip attempts to guess the name for the decompressed
file from that of the compressed file as follows:

@multitable {anyothername} {becomes} {anyothername.out}
@item filename.lz  @tab becomes @tab filename
@item filename.tlz @tab becomes @tab filename.tar
@item anyothername @tab becomes @tab anyothername.out
@end multitable

(De)compressing a file is much like copying or moving it; therefore plzip
preserves the access and modification dates, permissions, and, when
possible, ownership of the file just as "cp -p" does. (If the user ID or
the group ID can't be duplicated, the file permission bits S_ISUID and
S_ISGID are cleared).

Plzip is able to read from some types of non regular files if the
@samp{--stdout} option is specified.

If no file names are specified, plzip compresses (or decompresses) from
standard input to standard output. In this case, plzip will decline to
write compressed output to a terminal, as this would be entirely
incomprehensible and therefore pointless.

Plzip will correctly decompress a file which is the concatenation of two
or more compressed files. The result is the concatenation of the
corresponding uncompressed files. Integrity testing of concatenated
compressed files is also supported.

WARNING! Even if plzip is bug-free, other causes may result in a corrupt
compressed file (bugs in the system libraries, memory errors, etc).
Therefore, if the data you are going to compress are important, give the
@samp{--keep} option to plzip and do not remove the original file until
you verify the compressed file with a command like
@w{@samp{plzip -cd file.lz | cmp file -}}.


@node Invoking plzip
@chapter Invoking plzip
@cindex invoking
@cindex options
@cindex usage
@cindex version

The format for running plzip is:

@example
plzip [@var{options}] [@var{files}]
@end example

Plzip supports the following options:

@table @code
@item -h
@itemx --help
Print an informative help message describing the options and exit.

@item -V
@itemx --version
Print the version number of plzip on the standard output and exit.

@anchor{--data-size}
@item -B @var{bytes}
@itemx --data-size=@var{bytes}
Set the size of the input data blocks, in bytes. The input file will be
divided in chunks of this size before compression is performed. Valid
values range from 8 KiB to 1 GiB. Default value is two times the
dictionary size, except for option @samp{-0} where it defaults to 1 MiB.
Plzip will reduce the dictionary size if it is larger than the chosen
data size.

@item -c
@itemx --stdout
Compress or decompress to standard output. Needed when reading from a
named pipe (fifo) or from a device.

@item -d
@itemx --decompress
Decompress.

@item -f
@itemx --force
Force overwrite of output files.

@item -F
@itemx --recompress
Force re-compression of files whose name already has the @samp{.lz} or
@samp{.tlz} suffix.

@item -k
@itemx --keep
Keep (don't delete) input files during compression or decompression.

@item -m @var{bytes}
@itemx --match-length=@var{bytes}
Set the match length limit in bytes. After a match this long is found,
the search is finished. Valid values range from 5 to 273. Larger values
usually give better compression ratios but longer compression times.

@item -n @var{n}
@itemx --threads=@var{n}
Set the number of worker threads. Valid values range from 1 to "as many
as your system can support". If this option is not used, plzip tries to
detect the number of processors in the system and use it as default
value. @w{@samp{plzip --help}} shows the system's default value.

Note that the number of usable threads is limited to @w{ceil( file_size
/ data_size )} during compression (@pxref{Minimum file sizes}), and to
the number of members in the input during decompression.

@item -o @var{file}
@itemx --output=@var{file}
When reading from standard input and @samp{--stdout} has not been
specified, use @samp{@var{file}} as the virtual name of the uncompressed
file. This produces a file named @samp{@var{file}} when decompressing,
and a file named @samp{@var{file}.lz} when compressing.

@item -q
@itemx --quiet
Quiet operation. Suppress all messages.

@item -s @var{bytes}
@itemx --dictionary-size=@var{bytes}
Set the dictionary size limit in bytes. Valid values range from 4 KiB to
512 MiB. Plzip will use the smallest possible dictionary size for each
file without exceeding this limit. 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 (@var{bytes} / 16) to it.

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.

@item -t
@itemx --test
Check integrity of the specified file(s), but don't decompress them.
This really performs a trial decompression and throws away the result.
Use it together with @samp{-v} to see information about the file.

@item -v
@itemx --verbose
Verbose mode.@*
When compressing, show the compression ratio for each file processed. A
second @samp{-v} shows the progress of compression.@*
When decompressing or testing, further -v's (up to 4) increase the
verbosity level, showing status, compression ratio, dictionary size,
decompressed size, and compressed size.

@item -0 .. -9
Set the compression parameters (dictionary size and match length limit)
as shown in the table below. Note that @samp{-9} can be much slower than
@samp{-0}. These options have no effect when decompressing.

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 @samp{--match-length} and
@samp{--dictionary-size} options directly to achieve optimal
performance.

@multitable {Level} {Dictionary size} {Match length limit}
@item Level @tab Dictionary size @tab Match length limit
@item -0 @tab 64 KiB @tab  16 bytes
@item -1 @tab  1 MiB @tab   5 bytes
@item -2 @tab  1.5 MiB @tab   6 bytes
@item -3 @tab  2 MiB @tab   8 bytes
@item -4 @tab  3 MiB @tab  12 bytes
@item -5 @tab  4 MiB @tab  20 bytes
@item -6 @tab  8 MiB @tab  36 bytes
@item -7 @tab 16 MiB @tab  68 bytes
@item -8 @tab 24 MiB @tab 132 bytes
@item -9 @tab 32 MiB @tab 273 bytes
@end multitable

@item --fast
@itemx --best
Aliases for GNU gzip compatibility.

@end table

Numbers given as arguments to options may be followed by a multiplier
and an optional @samp{B} for "byte".

Table of SI and binary prefixes (unit multipliers):

@multitable {Prefix} {kilobyte  (10^3 = 1000)} {|} {Prefix} {kibibyte (2^10 = 1024)}
@item Prefix @tab Value               @tab | @tab Prefix @tab Value
@item k @tab kilobyte  (10^3 = 1000)  @tab | @tab Ki @tab kibibyte (2^10 = 1024)
@item M @tab megabyte  (10^6)         @tab | @tab Mi @tab mebibyte (2^20)
@item G @tab gigabyte  (10^9)         @tab | @tab Gi @tab gibibyte (2^30)
@item T @tab terabyte  (10^12)        @tab | @tab Ti @tab tebibyte (2^40)
@item P @tab petabyte  (10^15)        @tab | @tab Pi @tab pebibyte (2^50)
@item E @tab exabyte   (10^18)        @tab | @tab Ei @tab exbibyte (2^60)
@item Z @tab zettabyte (10^21)        @tab | @tab Zi @tab zebibyte (2^70)
@item Y @tab yottabyte (10^24)        @tab | @tab Yi @tab yobibyte (2^80)
@end multitable

@sp 1
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 plzip to panic.


@node Program design
@chapter Program design
@cindex program design

When compressing, plzip divides the input file into chunks and
compresses as many chunks simultaneously as worker threads are chosen,
creating a multi-member compressed file.

When decompressing, plzip decompresses as many members simultaneously as
worker threads are chosen. Files that were compressed with lzip will not
be decompressed faster than using lzip (unless the @samp{-b} option was
used) because lzip usually produces single-member files, which can't be
decompressed in parallel.

For each input file, a splitter thread and several worker threads are
created, acting the main thread as muxer (multiplexer) thread. A "packet
courier" takes care of data transfers among threads and limits the
maximum number of data blocks (packets) being processed simultaneously.

The splitter reads data blocks from the input file, and distributes them
to the workers. The workers (de)compress the blocks received from the
splitter. The muxer collects processed packets from the workers, and
writes them to the output file.

When decompressing from a regular file, the splitter is removed and the
workers read directly from the input file. If the output file is also a
regular file, the muxer is also removed and the workers write directly
to the output file. With these optimizations, the use of RAM is greatly
reduced and the decompression speed of large files with many members is
only limited by the number of processors available and by I/O speed.


@node File format
@chapter File format
@cindex file format

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

@sp 1
In the diagram below, a box like this:
@verbatim
+---+
|   | <-- the vertical bars might be missing
+---+
@end verbatim

represents one byte; a box like this:
@verbatim
+==============+
|              |
+==============+
@end verbatim

represents a variable number of bytes.

@sp 1
A lzip file consists of a series of "members" (compressed data sets).
The members simply appear one after another in the file, with no
additional information before, between, or after them.

Each member has the following structure:
@verbatim
+--+--+--+--+----+----+=============+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| ID string | VN | DS | Lzma stream | CRC32 |   Data size   |  Member size  |
+--+--+--+--+----+----+=============+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
@end verbatim

All multibyte values are stored in little endian order.

@table @samp
@item ID string
A four byte string, identifying the lzip format, with the value "LZIP"
(0x4C, 0x5A, 0x49, 0x50).

@item VN (version number, 1 byte)
Just in case something needs to be modified in the future. 1 for now.

@item DS (coded dictionary size, 1 byte)
The dictionary size is calculated by taking a power of 2 (the base size)
and substracting from it a fraction between 0/16 and 7/16 of the base
size.@*
Bits 4-0 contain the base 2 logarithm of the base size (12 to 29).@*
Bits 7-5 contain the numerator of the fraction (0 to 7) to substract
from the base size to obtain the dictionary size.@*
Example: 0xD3 = 2^19 - 6 * 2^15 = 512 KiB - 6 * 32 KiB = 320 KiB@*
Valid values for dictionary size range from 4 KiB to 512 MiB.

@item Lzma stream
The lzma stream, finished by an end of stream marker. Uses default
values for encoder properties.
@ifnothtml
@xref{Stream format,,,lzip},
@end ifnothtml
@ifhtml
See
@uref{http://www.nongnu.org/lzip/manual/lzip_manual.html#Stream-format,,Stream format}
@end ifhtml
for a complete description.

@item CRC32 (4 bytes)
CRC of the uncompressed original data.

@item Data size (8 bytes)
Size of the uncompressed original data.

@item Member size (8 bytes)
Total size of the member, including header and trailer. This field acts
as a distributed index, allows the verification of stream integrity, and
facilitates safe recovery of undamaged members from multi-member files.

@end table


@node Memory requirements
@chapter Memory required to compress and decompress
@cindex memory requirements

The amount of memory required @strong{per thread} is approximately the
following:

@itemize @bullet
@item
For compression at level -0; 1.5 MiB plus 3 times the data size
(@pxref{--data-size}). Default is 4.5 MiB.

@item
For compression at other levels; 11 times the dictionary size plus 3
times the data size. Default is 136 MiB.

@item
For decompression of a regular (seekable) file to another regular file,
or for testing of a regular file; the dictionary size.

(Note that regular files with more than 1024 bytes of trailing garbage
are treated as non-seekable).

@item
For testing of a non-seekable file or of standard input; the dictionary
size plus up to 5 MiB.

@item
For decompression of a regular file to a non-seekable file or to
standard output; the dictionary size plus up to 32 MiB.

@item
For decompression of a non-seekable file or of standard input; the
dictionary size plus up to 35 MiB.
@end itemize


@node Minimum file sizes
@chapter Minimum file sizes required for full compression speed
@cindex minimum file sizes

When compressing, plzip divides the input file into chunks and
compresses as many chunks simultaneously as worker threads are chosen,
creating a multi-member compressed file.

For this to work as expected (and roughly multiply the compression speed
by the number of available processors), the uncompressed file must be at
least as large as the number of worker threads times the chunk size
(@pxref{--data-size}). Else some processors will not get any data to
compress, and compression will be proportionally slower. The maximum
speed increase achievable on a given file is limited by the ratio
@w{(file_size / data_size)}.

The following table shows the minimum uncompressed file size needed for
full use of N processors at a given compression level, using the default
data size for each level:

@multitable {Processors} {512 MiB} {512 MiB} {512 MiB} {512 MiB} {512 MiB} {512 MiB}
@headitem Processors @tab 2 @tab 4 @tab 8 @tab 16 @tab 64 @tab 256
@item Level
@item -0 @tab   2 MiB @tab   4 MiB @tab   8 MiB @tab  16 MiB @tab  64 MiB @tab 256 MiB
@item -1 @tab   4 MiB @tab   8 MiB @tab  16 MiB @tab  32 MiB @tab 128 MiB @tab 512 MiB
@item -2 @tab   6 MiB @tab  12 MiB @tab  24 MiB @tab  48 MiB @tab 192 MiB @tab 768 MiB
@item -3 @tab   8 MiB @tab  16 MiB @tab  32 MiB @tab  64 MiB @tab 256 MiB @tab   1 GiB
@item -4 @tab  12 MiB @tab  24 MiB @tab  48 MiB @tab  96 MiB @tab 384 MiB @tab 1.5 GiB
@item -5 @tab  16 MiB @tab  32 MiB @tab  64 MiB @tab 128 MiB @tab 512 MiB @tab   2 GiB
@item -6 @tab  32 MiB @tab  64 MiB @tab 128 MiB @tab 256 MiB @tab   1 GiB @tab   4 GiB
@item -7 @tab  64 MiB @tab 128 MiB @tab 256 MiB @tab 512 MiB @tab   2 GiB @tab   8 GiB
@item -8 @tab  96 MiB @tab 192 MiB @tab 384 MiB @tab 768 MiB @tab   3 GiB @tab  12 GiB
@item -9 @tab 128 MiB @tab 256 MiB @tab 512 MiB @tab   1 GiB @tab   4 GiB @tab  16 GiB
@end multitable


@node Problems
@chapter Reporting bugs
@cindex bugs
@cindex getting help

There are probably bugs in plzip. 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 plzip, please send electronic mail to
@email{lzip-bug@@nongnu.org}. Include the version number, which you can
find by running @w{@code{plzip --version}}.


@node Concept index
@unnumbered Concept index

@printindex cp

@bye