[issue23200] Deprecate the zlib decompressor’s flush() method

Martin Panter added the comment: New version of the patch merged with current code and cleaning up the zipfile EOF detection. -- Added file: http://bugs.python.org/file41780/deprecate-flush.v2.patch ___ Python tracker

[issue23200] Deprecate the zlib decompressor’s flush() method

Serhiy Storchaka added the comment: After reading zlib manual it looks to me that your are right, flush() has no other effect except a little less memory consumption. This can be not important in the context of Python. Hope Nadeem will confirm this. -- assignee: docs@python -> nadeem.v

[issue23200] Deprecate the zlib decompressor’s flush() method

Martin Panter added the comment: And regarding other internal buffers and state, calling flush() or decompress() guarantees you get as much data as possible, but there is no error checking. You have to check the eof attribute to ensure everything is done, otherwise I don’t think there is a way

[issue23200] Deprecate the zlib decompressor’s flush() method

Martin Panter added the comment: For the bz2 and lzma modules, neither decompressor classes have a flush() method, only the compressors. -- ___ Python tracker ___ __

[issue23200] Deprecate the zlib decompressor’s flush() method

Serhiy Storchaka added the comment: Besides unconsumed_tail there is an internal buffer. Even if flush() is no longer needed in the zlib decompresser (I don't know), I doubt about bz2 and lzma. -- nosy: +serhiy.storchaka ___ Python tracker

[issue23200] Deprecate the zlib decompressor’s flush() method

Martin Panter added the comment: See Issue 224981 (bug) and Issue 403373 (patch) about the Z_SYNC_FLUSH change (the numbers in the commit message are different and do not work). I committed my doc change to 2.7 and 3.4+, which leaves the main problem of what to do about flush(). I propose depr