zstd.1: added advanced compression options

This commit is contained in:
Przemyslaw Skibinski 2016-12-14 16:10:13 +01:00
parent c71e552b2e
commit 25314428c9

View File

@ -264,6 +264,115 @@ and weight typically 100x the target dictionary size (for example, 10 MB for a 1
cut file into independent blocks of size # (default: no block)
.SH ADVANCED COMPRESSION OPTIONS
.TP
.B \--zstd[=\fIoptions\fR]
.PD
\fBzstd\fR provides 22 predefined compression levels. The selected or default predefined compression level can be changed with advanced compression options.
The \fIoptions\fR are provided as a comma-separated list. You may specify only the \fIoptions\fR you want to change and the rest will be taken from the selected or default compression level.
The list of available \fIoptions\fR:
.RS
.TP
.BI strategy= strat
.PD 0
.TP
.BI strat= strat
.PD
Specify a strategy used by a match finder.
.IP ""
There are 8 strategies numbered from 0 to 7, from faster to stronger:
0=ZSTD_fast, 1=ZSTD_dfast, 2=ZSTD_greedy, 3=ZSTD_lazy, 4=ZSTD_lazy2, 5=ZSTD_btlazy2, 6=ZSTD_btopt, 7=ZSTD_btopt2.
.IP ""
.TP
.BI windowLog= wlog
.PD 0
.TP
.BI wlog= wlog
.PD
Specify the maximum number of bits for a match distance.
.IP ""
The higher number of bits increases the chance to find a match what usually improves compression ratio.
It also increases memory requirements for compressor and decompressor.
.IP ""
The minimum \fIwlog\fR is 10 (1 KiB) and the maximum is 25 (32 MiB) for 32-bit compilation and 27 (128 MiB) for 64-bit compilation.
.IP ""
.TP
.BI hashLog= hlog
.PD 0
.TP
.BI hlog= hlog
.PD
Specify the maximum number of bits for a hash table.
.IP ""
The bigger hash table causes less collisions what usually make compression faster but requires more memory during compression.
.IP ""
The minimum \fIhlog\fR is 6 (64 B) and the maximum is 25 (32 MiB) for 32-bit compilation and 27 (128 MiB) for 64-bit compilation.
.TP
.BI chainLog= clog
.PD 0
.TP
.BI clog= clog
.PD
Specify the maximum number of bits for a hash chain or a binary tree.
.IP ""
The higher number of bits increases the chance to find a match what usually improves compression ratio.
It also slows down compression speed and increases memory requirements for compression.
This option is ignored for the ZSTD_fast strategy.
.IP ""
The minimum \fIclog\fR is 6 (64 B) and the maximum is 26 (64 MiB) for 32-bit compilation and 28 (256 MiB) for 64-bit compilation.
.IP ""
.TP
.BI searchLog= slog
.PD 0
.TP
.BI slog= slog
.PD
Specify the maximum number of searches in a hash chain or a binary tree using logarithmic scale.
.IP ""
The bigger number of searches increases the chance to find a match what usually improves compression ratio but decreases compression speed.
.IP ""
The minimum \fIslog\fR is 1 and the maximum is 24 for 32-bit compilation and 26 for 64-bit compilation.
.IP ""
.TP
.BI searchLength= slen
.PD 0
.TP
.BI slen= slen
.PD
Specify the minimum searched length of a match in a hash table.
.IP ""
The bigger search length usually decreases compression ratio but improves decompression speed.
.IP ""
The minimum \fIslen\fR is 3 and the maximum is 7.
.IP ""
.TP
.BI targetLength= tlen
.PD 0
.TP
.BI tlen= tlen
.PD
Specify the minimum match length that causes a match finder to interrupt searching of better matches.
.IP ""
The bigger minimum match length usually improves compression ratio but decreases compression speed.
This option is used only with ZSTD_btopt and ZSTD_btopt2 strategies.
.IP ""
The minimum \fItlen\fR is 4 and the maximum is 999.
.IP ""
.PP
.B An example
.br
The following parameters sets advanced compression options to predefined level 19 for files bigger than 256 KB:
.IP ""
\fB--zstd=\fRwindowLog=23,chainLog=23,hashLog=22,searchLog=6,searchLength=3,targetLength=48,strategy=6
.SH BUGS
Report bugs at:- https://github.com/facebook/zstd/issues