Changelog¶
Version 0.3.4¶
Drop Python 3.6 support.
Version 0.3.3¶
Drop Python 3.5 support and tag Python 3.9 and 3.10 support.
Version 0.3.2¶
Fix failing test_write_csv_hash()
on big-endian platforms.
Version 0.3.1¶
Add workaround for PyPy to write_csv()
when used with a hashlib
hash.
Version 0.3¶
Add autocompress
option to read_csv()
and write_csv()
(Python 3 only): automatically (de)compresses if file is a path and ends with
'.bz2'
, '.gz'
, or '.xz'
.
Version 0.2¶
Add read_csv()
and write_csv()
functions for working with open
(e.g. binary) streams (file-like objects). Available under Python 3 only.
Version 0.1.6¶
Tag Python 3.8 support.
Version 0.1.5¶
Add row_name
argument to NamedTupleReader
.
Drop Python 3.4 support, tag Python 3.7 support.
Extend documentation and improve formatting.
Version 0.1.4¶
Document workarounds for https://bugs.python.org/issue12178 and https://bugs.python.org/issue31590 bugs.
Add rowtype='namedtuple'
for collections.namedtuple
reader and writer.
Version 0.1.3¶
Fix missing mock
install requirement in setup.py
.
Version 0.1.2¶
Convert unicode values for format parameters like delimiter
to bytes on PY2.
Add PY2 backport of csv.unix_dialect
.
Add QUOTE_*
constants, Error
, dialects, and associated functions from csv
to csv23 namespace.
Version 0.1.1¶
Fix issue31590()
check.
Add documentation and CI links to README.txt
.
Version 0.1¶
Initial release (provisional API).