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

Fix an error when converting invalid html with empty tags to text

Browse Source
This commit is contained in:
Kovid Goyal 2018-10-30 09:33:21 +05:30
parent 0626088861
commit 8d70476517
No known key found for this signature in database
GPG Key ID: 06BC317B515ACE7C
1 changed files with 5 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
src/calibre/utils/html2text.py
View File

@ -50,6 +50,7 @@ def name2cp(k):
return int(k[2:-1]) # not in latin-1 return int(k[2:-1]) # not in latin-1
return ord(codecs.latin_1_decode(k)[0]) return ord(codecs.latin_1_decode(k)[0])
unifiable = {'rsquo':"'", 'lsquo':"'", 'rdquo':'"', 'ldquo':'"', unifiable = {'rsquo':"'", 'lsquo':"'", 'rdquo':'"', 'ldquo':'"',
'copy':'(C)', 'mdash':'--', 'nbsp':' ', 'rarr':'->', 'larr':'<-', 'middot':'*', 'copy':'(C)', 'mdash':'--', 'nbsp':' ', 'rarr':'->', 'larr':'<-', 'middot':'*',
'ndash':'-', 'oelig':'oe', 'aelig':'ae', 'ndash':'-', 'oelig':'oe', 'aelig':'ae',
@ -96,6 +97,7 @@ def replaceEntities(s):
else: else:
return entityref(s) return entityref(s)
r_unescape = re.compile(r"&(#?[xX]?(?:[0-9a-fA-F]+|\w{1,8}));") r_unescape = re.compile(r"&(#?[xX]?(?:[0-9a-fA-F]+|\w{1,8}));")
@ -150,7 +152,7 @@ def optwrap(text):
def hn(tag): def hn(tag):
if tag[0] == 'h' and len(tag) == 2: if tag and tag[0] == 'h' and len(tag) == 2:
try: try:
n = int(tag[1]) n = int(tag[1])
if n in range(1, 10): if n in range(1, 10):
@ -364,7 +366,7 @@ class _html2text(sgmllib.SGMLParser):
if not self.quiet: if not self.quiet:
if puredata and not self.pre: if puredata and not self.pre:
data = re.sub('\s+', ' ', data) data = re.sub(r'\s+', ' ', data)
if data and data[0] == ' ': if data and data[0] == ' ':
self.space = 1 self.space = 1
data = data[1:] data = data[1:]
@ -435,6 +437,7 @@ def html2text_file(html, out=wrapwrite, baseurl=''):
def html2text(html, baseurl=''): def html2text(html, baseurl=''):
return optwrap(html2text_file(html, None, baseurl)) return optwrap(html2text_file(html, None, baseurl))
if __name__ == "__main__": if __name__ == "__main__":
baseurl = '' baseurl = ''
if sys.argv[1:]: if sys.argv[1:]:
@ -461,4 +464,3 @@ if __name__ == "__main__":
else: else:
data = sys.stdin.read().decode('utf8') data = sys.stdin.read().decode('utf8')
wrapwrite(html2text(data, baseurl)) wrapwrite(html2text(data, baseurl))