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

Automatic detection of BD files.

Browse Source
This commit is contained in:
Kovid Goyal 2007-10-10 18:46:07 +00:00
parent 6cfd3561e6
commit d8b4bffb7e
1 changed files with 8 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
src/libprs500/ebooks/lrf/html/convert_from.py
View File

@ -228,6 +228,9 @@ class HTMLConverter(object):
def is_baen(self, soup): def is_baen(self, soup):
return bool(soup.find('meta', attrs={'name':'Publisher', return bool(soup.find('meta', attrs={'name':'Publisher',
'content':re.compile('Baen', re.IGNORECASE)})) 'content':re.compile('Baen', re.IGNORECASE)}))
def is_book_designer(self, soup):
return bool(soup.find('h2', attrs={'id':'BookTitle'}))
def preprocess(self, raw): def preprocess(self, raw):
nmassage = copy.copy(BeautifulSoup.MARKUP_MASSAGE) nmassage = copy.copy(BeautifulSoup.MARKUP_MASSAGE)
@ -254,6 +257,10 @@ class HTMLConverter(object):
self.baen = True self.baen = True
self.logger.info('Baen file detected. Re-parsing...') self.logger.info('Baen file detected. Re-parsing...')
return self.preprocess(raw) return self.preprocess(raw)
if not self.book_designer and self.is_book_designer(soup):
self.book_designer = True
self.logger.info('Book Designer file detected. Re-parsing...')
return self.preprocess(raw)
if self.book_designer: if self.book_designer:
t = soup.find(id='BookTitle') t = soup.find(id='BookTitle')
if t: if t:
@ -268,7 +275,7 @@ class HTMLConverter(object):
self.logger.info('Written preprocessed HTML to '+dump.name) self.logger.info('Written preprocessed HTML to '+dump.name)
dump.close() dump.close()
print soup #print soup
return soup return soup
def start_on_file(self, path, is_root=True, link_level=0): def start_on_file(self, path, is_root=True, link_level=0):