You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
hz is a Simplified Chinese codec, available in Python since around 2004.
However, hz encoder has a serious bug, it forgets to escape ~
>>> 'hi~'.encode('hz')
b'hi~' # the correct output should be b'hi~~'
As a result, we can't finish a roundtrip:
>>> b'hi~'.decode('hz')
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
UnicodeDecodeError: 'hz' codec can't decode byte 0x7e in position 2: incomplete multibyte
In these years, no one has reported this bug, so I think it's pretty safe to remove hz codec.
FYI:
HZ codec is a 7-bit wrapper for GB2312, was formerly commonly used in email and USENET postings. It was designed in 1989 by Fung Fung Lee, and subsequently codified in 1995 into RFC 1843.
It was popular in USENET networks, which in the late 1980s and early 1990s, generally did not allow transmission of 8-bit characters or escape characters.
But later, I thought it's a good opportunity to remove this codec, this serious bug indicates that almost no one is using it. But fixing will create a possibility that someone will using it in future.
So I suggest we don't fix it, just remove it or leave it as is.
hz is outdated, searching on internet almost no one talking about it.
Or do you know other bugs?
It has another small bug in decoder, about state switch, but it's trivial, also fixed in the patch.
We seldom just remove things; we usually deprecate in the doc and if possible, issue a runtime warning.
This is probably not the only obsolete codec. There should be a uniform policy for deprecation and removal, if ever. But for any codec, there might be archives, even if the codec is not used for new files.
If the codec is buggy, I think it should be fixed. Bt you yourself closed bpo-24117, suggesting that you did not believe that the patches should be applied.
Activity
animalize commentedon Apr 6, 2017
hz is a Simplified Chinese codec, available in Python since around 2004.
In these years, no one has reported this bug, so I think it's pretty safe to remove hz codec.
FYI:
HZ codec is a 7-bit wrapper for GB2312, was formerly commonly used in email and USENET postings. It was designed in 1989 by Fung Fung Lee, and subsequently codified in 1995 into RFC 1843.
It was popular in USENET networks, which in the late 1980s and early 1990s, generally did not allow transmission of 8-bit characters or escape characters.
https://en.wikipedia.org/wiki/HZ_(character_encoding)
Does other languages have hz codec?
Java 8: no [1]
.NET: yes [2]
PHP: yes [3]
Perl: yes [4]
[1] http://docs.oracle.com/javase/8/docs/technotes/guides/intl/encoding.doc.html
[2] https://msdn.microsoft.com/en-us/library/system.text.encoding(v=vs.110).aspx
[3] http://php.net/manual/en/mbstring.supported-encodings.php
[4] http://perldoc.perl.org/Encode/CN.html
vstinner commentedon Apr 6, 2017
Can't we fix the bug instead of removing the whole codec? Or do you know other bugs?
The bug is only on the encoder part, right? I see unit test for '~' on the hz decoder.
animalize commentedon Apr 6, 2017
I tried to fix this two years ago, here is the patch (not merged):
http://bugs.python.org/review/24117/diff/14803/Modules/cjkcodecs/_codecs_cn.c
But later, I thought it's a good opportunity to remove this codec, this serious bug indicates that almost no one is using it. But fixing will create a possibility that someone will using it in future.
So I suggest we don't fix it, just remove it or leave it as is.
hz is outdated, searching on internet almost no one talking about it.
terryjreedy commentedon Apr 7, 2017
We seldom just remove things; we usually deprecate in the doc and if possible, issue a runtime warning.
This is probably not the only obsolete codec. There should be a uniform policy for deprecation and removal, if ever. But for any codec, there might be archives, even if the codec is not used for new files.
If the codec is buggy, I think it should be fixed. Bt you yourself closed bpo-24117, suggesting that you did not believe that the patches should be applied.
vstinner commentedon Apr 8, 2017
"But for any codec, there might be archives, even if the codec is not
used for new files."
The bug is in the encoder. The codec is still usable to *decode*
files. So maybe a few people use it but didn't notice the encoder bug?
animalize commentedon Apr 8, 2017
From my subjective feelings, probably no old archives still exist, but I can't assert it. That's why I suggest remove it, or at least don't fix it.
Ah, let's slow down the pace, this bug exists over a dacade, we don't need to solve it at once.
I closed bpo-24117, because it became a soup of small issues, so I split it into individual issues (such as this issue).
[-]Remove hz codec[/-][+]Fix handling escape characters in HZ codec[/+]zhangyangyu commentedon May 22, 2017
New changeset 89a5e03 by Xiang Zhang in branch 'master':
bpo-30003: Fix handling escape characters in HZ codec (bpo-1556)
89a5e03
zhangyangyu commentedon May 22, 2017
New changeset 65440f8 by Xiang Zhang in branch '3.5':
bpo-30003: Fix handling escape characters in HZ codec (bpo-1556) (bpo-1718)
65440f8
zhangyangyu commentedon May 22, 2017
New changeset 54af41d by Xiang Zhang in branch '3.6':
bpo-30003: Fix handling escape characters in HZ codec (bpo-1556) (bpo-1719)
54af41d
zhangyangyu commentedon May 22, 2017
New changeset 6e1b832 by Xiang Zhang in branch '2.7':
bpo-30003: Fix handling escape characters in HZ codec (bpo-1720) (bpo-1556)
6e1b832