Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-09 03:04:10 -04:00
Code Issues Packages Projects Releases Wiki Activity

DOCX Input: Fix right-to-left text support only working for block level styles, not character level styles.

Browse Source
This commit is contained in:
Kovid Goyal 2016-08-19 23:52:28 +05:30
parent fefb3ee882
commit 7cb8a5f4d2
2 changed files with 22 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

22
src/calibre/ebooks/docx/cleanup.py
View File

@ -114,7 +114,7 @@ def cleanup_markup(log, root, styles, dest_dir, detect_cover, XPath):
# Merge consecutive spans that have the same styling
current_run = []
for span in root.xpath('//span[not(@style or @lang)]'):
for span in root.xpath('//span[not(@style or @lang or @dir)]'):
if not current_run:
current_run.append(span)
else:
@ -130,6 +130,22 @@ def cleanup_markup(log, root, styles, dest_dir, detect_cover, XPath):
# element
class_map = dict(styles.classes.itervalues())
parents = ('p', 'div') + tuple('h%d' % i for i in xrange(1, 7))
for parent in root.xpath('//*[(%s)]' % ' or '.join('name()="%s"' % t for t in parents)):
# If all spans have dir="rtl" and the parent does not have dir set,
# move the dir to the parent.
if len(parent) and (parent.get('dir') or 'rtl') == 'rtl':
has_rtl_children = False
for child in parent.iterchildren('span'):
if child.get('dir') == 'rtl':
has_rtl_children = True
else:
has_rtl_children = False
break
if has_rtl_children:
parent.set('dir', 'rtl')
for child in parent.iterchildren():
del child.attrib['dir']
for parent in root.xpath('//*[(%s) and count(span)=1]' % ' or '.join('name()="%s"' % t for t in parents)):
if len(parent) == 1 and not parent.text and not parent[0].tail and not parent[0].get('id', None):
# We have a block whose contents are entirely enclosed in a <span>
@ -145,6 +161,8 @@ def cleanup_markup(log, root, styles, dest_dir, detect_cover, XPath):
parent.remove(span)
if span.get('lang'):
parent.set('lang', span.get('lang'))
if span.get('dir'):
parent.set('dir', span.get('dir'))
for child in span:
parent.append(child)
@ -160,7 +178,7 @@ def cleanup_markup(log, root, styles, dest_dir, detect_cover, XPath):
del span.attrib['class']
# Get rid of <span>s that have no styling
for span in root.xpath('//span[not(@class or @id or @style or @lang)]'):
for span in root.xpath('//span[not(@class or @id or @style or @lang or @dir)]'):
lift(span)
# Convert <p><br style="page-break-after:always"> </p> style page breaks

2
src/calibre/ebooks/docx/to_html.py
View File

@ -674,6 +674,8 @@ class Convert(object):
lang = html_lang(style.lang)
if lang is not None and lang != self.doc_lang:
ans.set('lang', lang)
if style.rtl is True:
ans.set('dir', 'rtl')
return ans
def add_frame(self, html_obj, style):