Add the "namereplace" error handler

[serhiy-storchaka]

BPO	19676
Nosy	@malemburg, @amauryfa, @ncoghlan, @vstinner, @ned-deily, @ezio-melotti, @stevendaprano, @ethanfurman, @serhiy-storchaka
Files	namereplace_errors.patch

^{Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.}

Show more details

GitHub fields:

assignee = 'https://github.com/serhiy-storchaka'
closed_at = <Date 2014-11-26.20:27:44.576>
created_at = <Date 2013-11-21.07:41:46.661>
labels = ['type-feature', 'expert-unicode']
title = 'Add the "namereplace" error handler'
updated_at = <Date 2014-11-26.20:27:44.575>
user = 'https://github.com/serhiy-storchaka'

bugs.python.org fields:

activity = <Date 2014-11-26.20:27:44.575>
actor = 'ned.deily'
assignee = 'serhiy.storchaka'
closed = True
closed_date = <Date 2014-11-26.20:27:44.576>
closer = 'ned.deily'
components = ['Unicode']
creation = <Date 2013-11-21.07:41:46.661>
creator = 'serhiy.storchaka'
dependencies = []
files = ['32748']
hgrepos = []
issue_num = 19676
keywords = ['patch', 'needs review']
message_count = 13.0
messages = ['203579', '203580', '231647', '231649', '231650', '231652', '231653', '231654', '231672', '231700', '231701', '231727', '231728']
nosy_count = 10.0
nosy_names = ['lemburg', 'amaury.forgeotdarc', 'ncoghlan', 'vstinner', 'ned.deily', 'ezio.melotti', 'steven.daprano', 'ethan.furman', 'python-dev', 'serhiy.storchaka']
pr_nums = []
priority = 'normal'
resolution = 'fixed'
stage = 'resolved'
status = 'closed'
superseder = None
type = 'enhancement'
url = 'https://bugs.python.org/issue19676'
versions = ['Python 3.5']

[serhiy-storchaka]

The proposed patch adds the "namereplace" error handler. This error handler is almost same as the "backslashreplace" error handler, but use \N{...} escape sequences if there is a character name in Unicode database. Result is a little more human-readable (but less portable) than with "backslashreplace".

>>> '∀ x∈ℜ'.encode('ascii', 'namereplace')
b'\\N{FOR ALL} x\\N{ELEMENT OF}\\N{BLACK-LETTER CAPITAL R}'

The proposition was discussed and bikeshedded on Python-Ideas: http://comments.gmane.org/gmane.comp.python.ideas/21296 .

[vstinner]

See also issue bpo-18234.

[serhiy-storchaka]

Ping.

[amauryfa]

The patch looks good to me.
But it seems that the reverse operation is not possible in the general case: .decode('unicode_escape') assumes a latin-1 or ascii encoding.
Should we document this?

[malemburg]

The patch looks good.

One nit: the name buffer length should be NAME_MAXLEN instead of 100.

[ncoghlan]

Patch looks good to me, too.

As far as Amaury's question goes, isn't the general reverse operation the same as for the existing backslashreplace handler?

That is, decode with the appropriate ASCII compatible encoding (since ASCII compatibility is needed for the escape sequences to be valid), then run the result through ast.literal_eval?

(I'll grant we don't currently provide guidance on reversing backslashreplace either, but addressing that sounds like a separate question from this change)

[python-dev]

New changeset 32d08aacffe0 by Serhiy Storchaka in branch 'default':
Issue bpo-19676: Added the "namereplace" error handler.
https://hg.python.org/cpython/rev/32d08aacffe0

[serhiy-storchaka]

Thank you all for reviews.

One nit: the name buffer length should be NAME_MAXLEN instead of 100.

NAME_MAXLEN is private name available only in Modules/unicodedata.c. Making it public name would be other issue. I have increased buffer size to 256.

[python-dev]

New changeset b6fab008d63a by Berker Peksag in branch 'default':
Issue bpo-19676: Tweak documentation a bit.
https://hg.python.org/cpython/rev/b6fab008d63a

[python-dev]

New changeset 21d1571c0533 by Serhiy Storchaka in branch 'default':
Issue bpo-19676: Fixed integer overflow issue in "namereplace" error handler.
https://hg.python.org/cpython/rev/21d1571c0533

[serhiy-storchaka]

Thank you Berker.