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

MOBI Input:Improve detection of Table of Contents. Fixes #1936 (SInce Calibre 0.4.134 medatada in reader doesn't works)

Browse Source
This commit is contained in:
Kovid Goyal 2009-02-26 14:13:43 -08:00
parent b5e9f19502
commit 4d44f2d8bd
1 changed files with 20 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

34
src/calibre/ebooks/mobi/reader.py
View File

@ -235,7 +235,7 @@ class MobiReader(object):
if self.verbose:
print 'Creating OPF...'
ncx = cStringIO.StringIO()
opf = self.create_opf(htmlfile, guide)
opf = self.create_opf(htmlfile, guide, root)
opf.render(open(os.path.splitext(htmlfile)[0]+'.opf', 'wb'), ncx)
ncx = ncx.getvalue()
if ncx:
@ -328,7 +328,7 @@ class MobiReader(object):
except ValueError:
pass
def create_opf(self, htmlfile, guide=None):
def create_opf(self, htmlfile, guide=None, root=None):
mi = self.book_header.exth.mi
opf = OPFCreator(os.path.dirname(htmlfile), mi)
if hasattr(self.book_header.exth, 'cover_offset'):
@ -347,21 +347,27 @@ class MobiReader(object):
if ref.type.lower() == 'toc':
toc = ref.href()
if toc:
index = self.processed_html.find('<a id="%s" name="%s"'%(toc.partition('#')[-1], toc.partition('#')[-1]))
elems = root.xpath('//*[@id="%s"]'%toc.partition('#')[-1])
tocobj = None
ent_pat = re.compile(r'&(\S+?);')
if index > -1:
raw = '<html><body>'+self.processed_html[index:]
root = html.fromstring(raw)
if elems:
tocobj = TOC()
for a in root.xpath('//a[@href]'):
try:
text = u' '.join([t.strip() for t in a.xpath('descendant::text()')])
except:
text = ''
text = ent_pat.sub(entity_to_unicode, text)
if a.get('href', '').startswith('#'):
tocobj.add_item(toc.partition('#')[0], a.attrib['href'][1:], text)
reached = False
for x in root.iter():
if x == elems[-1]:
reached = True
continue
if reached and x.tag == 'a':
href = x.get('href', '')
if href:
try:
text = u' '.join([t.strip() for t in \
x.xpath('descendant::text()')])
except:
text = ''
text = ent_pat.sub(entity_to_unicode, text)
tocobj.add_item(toc.partition('#')[0], href[1:],
text)
if tocobj is not None:
opf.set_toc(tocobj)