Linux Standard Base Specification 2.1 | ||
---|---|---|
<<< Previous | Next >>> |
The deflateSetDictionary() function shall initialize the compression dictionary associated with stream using the dictlen bytes referenced by dictionary.
The implementation may silently use a subset of the provided dictionary if the dictionary cannot fit in the current window associated with stream (see deflateInit2_()). The application should ensure that the dictionary is sorted such that the most commonly used strings occur at the end of the dictionary.
If the dictionary is successfully set, the Adler32 checksum of the entire
provided dictionary
shall be stored in the adler
member of
stream. This value may be used by the decompression
system to select the correct dictionary. The compression and decompression
systems must use the same dictionary.
stream shall reference an initialized compression
stream, with total_in
zero (i.e. no data
has been compressed since the stream was initialized).
On success, deflateSetDictionary() shall return Z_OK. Otherwise it shall return Z_STREAM_ERROR to indicate an error.
On error, deflateSetDictionary() shall return a value as described below:
Z_STREAM_ERROR
The state in stream is inconsistent, or
stream
was NULL
.
The application should provide a dictionary consisting of strings {{{ed note: do we really mean "strings"? Null terminated?}}} that are likely to be encountered in the data to be compressed. The application should ensure that the dictionary is sorted such that the most commonly used strings occur at the end of the dictionary.
The use of a dictionary is optional; however if the data to be compressed is relatively short and has a predictable structure, the use of a dictionary can substantially improve the compression ratio.
<<< Previous | Home | Next >>> |
deflateReset | Up | get_crc_table |