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

reordered chapter priority, added em-dashes to smarten punctuation

Browse Source
This commit is contained in:
ldolse 2011-01-03 01:11:34 -05:00
parent 7008e2a23a
commit a2e47dae8f
2 changed files with 4 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
src/calibre/ebooks/conversion/preprocess.py
View File

@ -353,7 +353,7 @@ class HTMLPreProcessor(object):
(re.compile(r'((?<=</a>)\s*file:////?[A-Z].*<br>|file:////?[A-Z].*<br>(?=\s*<hr>))', re.IGNORECASE), lambda match: ''),
# Center separator lines
(re.compile(u'<br>\s*(?P<break>([*#•]+\s*)+)\s*<br>'), lambda match: '<p>\n<p style="text-align:center">' + match.group(1) + '</p>'),
(re.compile(u'<br>\s*(?P<break>([*#•✦]+\s*)+)\s*<br>'), lambda match: '<p>\n<p style="text-align:center">' + match.group(1) + '</p>'),
# Remove page links
(re.compile(r'<a name=\d+></a>', re.IGNORECASE), lambda match: ''),
@ -567,6 +567,7 @@ class HTMLPreProcessor(object):
html = html.replace(stop, '-->')
# convert ellipsis to entities to prevent wrapping
html = re.sub('(?u)(?<=\w)\s?(\.\s?){2}\.', '&hellip;', html)
# nbsp gets changed to space: html = re.sub('(?u)(?<=\w)\s(\.\s?){2}\.', '&nbsp;&hellip;', html)
# convert double dashes to em-dash
html = re.sub('\s--\s', u'\u2014', html)
return substitute_entites(html)

2
src/calibre/ebooks/conversion/utils.py
View File

@ -156,9 +156,9 @@ class PreProcessor(object):
[r"[^'\"]?(Introduction|Synopsis|Acknowledgements|Chapter|Kapitel|Epilogue|Volume\s|Prologue|Book\s|Part\s|Dedication)\s*([\d\w-]+\:?\s*){0,4}", True, "Searching for common Chapter Headings"],
[r"[^'\"]?(\d+(\.|:)|CHAPTER)\s*([\dA-Z\-\'\"#,]+\s*){0,7}\s*", True, "Searching for numeric chapter headings"], # Numeric Chapters
[r"([A-Z]\s+){3,}\s*([\d\w-]+\s*){0,3}\s*", True, "Searching for letter spaced headings"], # Spaced Lettering
[r"[^'\"]?(\d+|CHAPTER)\s*([\dA-Z\-\'\"\?!#,]+\s*){0,7}\s*", True, "Searching for simple numeric chapter headings"], # Numeric Chapters, no dot or colon
[r"<b[^>]*>\s*(<span[^>]*>)?\s*(?!([*#•]+\s*)+)(\s*(?=[\w#\-*\s]+<)([\w#-*]+\s*){1,5}\s*)(</span>)?\s*</b>", True, "Searching for emphasized lines"], # Emphasized lines
[r"[^'\"]?(\d+\.?\s+([\d\w-]+\:?\'?-?\s?){0,5})\s*", True, "Searching for numeric chapters with titles"], # Numeric Titles
[r"[^'\"]?(\d+|CHAPTER)\s*([\dA-Z\-\'\"\?!#,]+\s*){0,7}\s*", True, "Searching for simple numeric chapter headings"], # Numeric Chapters, no dot or colon
[r"\s*[^'\"]?([A-Z#]+(\s|-){0,3}){1,5}\s*", False, "Searching for chapters with Uppercase Characters" ] # Uppercase Chapters
]