Archives/python-readability
2
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Corrects some things were 2to3 did to much.

Browse Source
This commit is contained in:
Martin Thurau 2015-04-29 19:33:43 +02:00
parent aa4132f57a
commit 3ac56329e2
2 changed files with 4 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
readability/htmls.py
View File

@ -36,7 +36,7 @@ def normalize_entities(cur_title):
u'\u00BB': '"',
u'"': '"',
}
for c, r in list(entities.items()):
for c, r in entities.items():
if c in cur_title:
cur_title = cur_title.replace(c, r)

6
readability/readability.py
View File

@ -89,7 +89,7 @@ def compile_pattern(elements):
else:
# assume string or string like object
elements = elements.split(',')
return re.compile('|'.join([re.escape(x.lower()) for x in elements]), re.U)
return re.compile(u'|'.join([re.escape(x.lower()) for x in elements]), re.U)
class Document:
"""Class to build a etree document out of html."""
@ -207,7 +207,7 @@ class Document:
# code never runs) which would require write this line as:
# write this line as
# Unparseable(str(e))
# but then we loose the traceback information. 3.4 on the
# but then we lose the traceback information. 3.4 on the
# other hand accepts the old syntax and would only complain
# at runtime.
raise Unparseable(str(e)), None, sys.exc_info()[2]
@ -262,7 +262,7 @@ class Document:
return output
def select_best_candidate(self, candidates):
sorted_candidates = sorted(list(candidates.values()), key=lambda x: x['content_score'], reverse=True)
sorted_candidates = sorted(candidates.values(), key=lambda x: x['content_score'], reverse=True)
for candidate in sorted_candidates[:5]:
elem = candidate['elem']
self.debug("Top 5 : %6.3f %s" % (