Closed GoogleCodeExporter closed 8 years ago
(with "buffer thats smaller than the chunkSize" i meant that the size of the
buffer is smaller than chunkSize after decompression)
Original comment by d...@bbcgmbh.de
on 1 Jan 2013 at 11:03
Well, imho, this is a correct behavior.
The function cannot decompress data, because it has not enough place to
decompress it.
So it returns an error, which is a negative value.
If it would return a positive value, then it would mean that it successfully
decoded the compressed stream, and the value would be the size of decoded data.
Original comment by yann.col...@gmail.com
on 1 Jan 2013 at 11:21
So i should just arbitrarily choose a large enough buffer size? I honestly
would expect to have the chance to increase the buffer size (for example gzip
offers this functionality). The function could just decompress the data until
the buffer is full and then request the buffer to be increased.
Original comment by d...@bbcgmbh.de
on 1 Jan 2013 at 11:25
What you seem to request will be part of the LZ4 streaming interface, which is
still a work in progress.
For the time being, the LZ4 "raw format" as delivered here is a block-oriented
algorithm. You are supposed to know the size of data to decode, or at least to
know the maximum size of data to decode. How this information is transmitted is
outside of the scope of LZ4 "raw format".
Original comment by yann.col...@gmail.com
on 2 Jan 2013 at 12:15
Closed. Will be part of the LZ4 streaming interface.
Original comment by yann.col...@gmail.com
on 3 Jan 2013 at 3:24
Original issue reported on code.google.com by
d...@bbcgmbh.de
on 1 Jan 2013 at 11:01