mirror of
https://github.com/python/cpython
synced 2024-09-15 22:10:19 +00:00
bpo-32174: Let .chm document display non-ASCII characters properly (GH-9758)
Let .chm document display non-ASCII characters properly Escape the `body` part of .chm source file to 7-bit ASCII, to fix visual effect on some MBCS Windows systems.
This commit is contained in:
parent
60d230c78f
commit
6261ae9b01
|
@ -14,7 +14,7 @@
|
|||
# ---------------------
|
||||
|
||||
extensions = ['sphinx.ext.coverage', 'sphinx.ext.doctest',
|
||||
'pyspecific', 'c_annotations']
|
||||
'pyspecific', 'c_annotations', 'escape4chm']
|
||||
|
||||
# General substitutions.
|
||||
project = 'Python'
|
||||
|
|
39
Doc/tools/extensions/escape4chm.py
Normal file
39
Doc/tools/extensions/escape4chm.py
Normal file
|
@ -0,0 +1,39 @@
|
|||
"""
|
||||
Escape the `body` part of .chm source file to 7-bit ASCII, to fix visual
|
||||
effect on some MBCS Windows systems.
|
||||
|
||||
https://bugs.python.org/issue32174
|
||||
"""
|
||||
|
||||
import re
|
||||
from html.entities import codepoint2name
|
||||
|
||||
# escape the characters which codepoint > 0x7F
|
||||
def _process(string):
|
||||
def escape(matchobj):
|
||||
codepoint = ord(matchobj.group(0))
|
||||
|
||||
name = codepoint2name.get(codepoint)
|
||||
if name is None:
|
||||
return '&#%d;' % codepoint
|
||||
else:
|
||||
return '&%s;' % name
|
||||
|
||||
return re.sub(r'[^\x00-\x7F]', escape, string)
|
||||
|
||||
def escape_for_chm(app, pagename, templatename, context, doctree):
|
||||
# only works for .chm output
|
||||
if not hasattr(app.builder, 'name') or app.builder.name != 'htmlhelp':
|
||||
return
|
||||
|
||||
# escape the `body` part to 7-bit ASCII
|
||||
body = context.get('body')
|
||||
if body is not None:
|
||||
context['body'] = _process(body)
|
||||
|
||||
def setup(app):
|
||||
# `html-page-context` event emitted when the HTML builder has
|
||||
# created a context dictionary to render a template with.
|
||||
app.connect('html-page-context', escape_for_chm)
|
||||
|
||||
return {'version': '1.0', 'parallel_read_safe': True}
|
|
@ -0,0 +1,2 @@
|
|||
chm document displays non-ASCII charaters properly on some MBCS Windows
|
||||
systems.
|
Loading…
Reference in a new issue