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

CHM Input: Fix handling of relative file paths in <img> tags. Fixes #7159 (Large CHM file fails conversion)

Browse Source
This commit is contained in:
Kovid Goyal 2010-10-17 09:05:40 -06:00
parent ca4953f028
commit 30b3d2a564
1 changed files with 30 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

42
src/calibre/ebooks/chm/reader.py
View File

@ -93,6 +93,7 @@ class CHMReader(CHMFile):
return data return data
def ExtractFiles(self, output_dir=os.getcwdu()): def ExtractFiles(self, output_dir=os.getcwdu()):
html_files = set([])
for path in self.Contents(): for path in self.Contents():
lpath = os.path.join(output_dir, path) lpath = os.path.join(output_dir, path)
self._ensure_dir(lpath) self._ensure_dir(lpath)
@ -106,14 +107,27 @@ class CHMReader(CHMFile):
lpath = lpath.split(';')[0] lpath = lpath.split(';')[0]
try: try:
with open(lpath, 'wb') as f: with open(lpath, 'wb') as f:
if guess_mimetype(path)[0] == ('text/html'):
data = self._reformat(data)
f.write(data) f.write(data)
try:
if 'html' in guess_mimetype(path)[0]:
html_files.add(lpath)
except:
pass
except: except:
if iswindows and len(lpath) > 250: if iswindows and len(lpath) > 250:
self.log.warn('%r filename too long, skipping'%path) self.log.warn('%r filename too long, skipping'%path)
continue continue
raise raise
for lpath in html_files:
with open(lpath, 'r+b') as f:
data = f.read()
data = self._reformat(data, lpath)
if isinstance(data, unicode):
data = data.encode('utf-8')
f.seek(0)
f.truncate()
f.write(data)
self._extracted = True self._extracted = True
files = [x for x in os.listdir(output_dir) if files = [x for x in os.listdir(output_dir) if
os.path.isfile(os.path.join(output_dir, x))] os.path.isfile(os.path.join(output_dir, x))]
@ -125,7 +139,7 @@ class CHMReader(CHMFile):
if self.hhc_path not in files and files: if self.hhc_path not in files and files:
self.hhc_path = files[0] self.hhc_path = files[0]
def _reformat(self, data): def _reformat(self, data, htmlpath):
try: try:
data = xml_to_unicode(data, strip_encoding_pats=True)[0] data = xml_to_unicode(data, strip_encoding_pats=True)[0]
soup = BeautifulSoup(data) soup = BeautifulSoup(data)
@ -169,15 +183,19 @@ class CHMReader(CHMFile):
br[0].extract() br[0].extract()
# some images seem to be broken in some chm's :/ # some images seem to be broken in some chm's :/
for img in soup('img'): base = os.path.dirname(htmlpath)
try: for img in soup('img', src=True):
# some are supposedly "relative"... lies. src = img['src']
while img['src'].startswith('../'): img['src'] = img['src'][3:] ipath = os.path.join(base, *src.split('/'))
# some have ";<junk>" at the end. if os.path.exists(ipath):
img['src'] = img['src'].split(';')[0] continue
except KeyError: src = src.split(';')[0]
# and some don't even have a src= ?! if not src: continue
pass ipath = os.path.join(base, *src.split('/'))
if not os.path.exists(ipath):
while src.startswith('../'):
src = src[3:]
img['src'] = src
try: try:
# if there is only a single table with a single element # if there is only a single table with a single element
# in the body, replace it by the contents of this single element # in the body, replace it by the contents of this single element