#13993: HTMLParser is now able to handle broken end tags.
diff --git a/Lib/HTMLParser.py b/Lib/HTMLParser.py
index 516bc70..6cc9ff1 100644
--- a/Lib/HTMLParser.py
+++ b/Lib/HTMLParser.py
@@ -23,6 +23,9 @@
piclose = re.compile('>')
commentclose = re.compile(r'--\s*>')
tagfind = re.compile('[a-zA-Z][-.a-zA-Z0-9:_]*')
+# see http://www.w3.org/TR/html5/tokenization.html#tag-open-state
+# and http://www.w3.org/TR/html5/tokenization.html#tag-name-state
+tagfind_tolerant = re.compile('[a-zA-Z][^\t\n\r\f />\x00]*')
attrfind = re.compile(
r'\s*((?<=[\'"\s])[^\s/>][^\s/=>]*)(\s*=+\s*'
@@ -243,7 +246,7 @@
# see http://www.w3.org/TR/html5/tokenization.html#bogus-comment-state
def parse_bogus_comment(self, i, report=1):
rawdata = self.rawdata
- if rawdata[i:i+2] != '<!':
+ if rawdata[i:i+2] not in ('<!', '</'):
self.error('unexpected call to parse_comment()')
pos = rawdata.find('>', i+2)
if pos == -1:
@@ -353,23 +356,38 @@
match = endendtag.search(rawdata, i+1) # >
if not match:
return -1
- j = match.end()
+ gtpos = match.end()
match = endtagfind.match(rawdata, i) # </ + tag + >
if not match:
if self.cdata_elem is not None:
- self.handle_data(rawdata[i:j])
- return j
- self.error("bad end tag: %r" % (rawdata[i:j],))
+ self.handle_data(rawdata[i:gtpos])
+ return gtpos
+ # find the name: w3.org/TR/html5/tokenization.html#tag-name-state
+ namematch = tagfind_tolerant.match(rawdata, i+2)
+ if not namematch:
+ # w3.org/TR/html5/tokenization.html#end-tag-open-state
+ if rawdata[i:i+3] == '</>':
+ return i+3
+ else:
+ return self.parse_bogus_comment(i)
+ tagname = namematch.group().lower()
+ # consume and ignore other stuff between the name and the >
+ # Note: this is not 100% correct, since we might have things like
+ # </tag attr=">">, but looking for > after tha name should cover
+ # most of the cases and is much simpler
+ gtpos = rawdata.find('>', namematch.end())
+ self.handle_endtag(tagname)
+ return gtpos+1
elem = match.group(1).lower() # script or style
if self.cdata_elem is not None:
if elem != self.cdata_elem:
- self.handle_data(rawdata[i:j])
- return j
+ self.handle_data(rawdata[i:gtpos])
+ return gtpos
self.handle_endtag(elem)
self.clear_cdata_mode()
- return j
+ return gtpos
# Overridable -- finish processing of start+end tag: <tag.../>
def handle_startendtag(self, tag, attrs):
diff --git a/Lib/test/test_htmlparser.py b/Lib/test/test_htmlparser.py
index 29a721c..c15bb66 100644
--- a/Lib/test/test_htmlparser.py
+++ b/Lib/test/test_htmlparser.py
@@ -202,12 +202,12 @@
self._run_check(["<!--abc-->", ""], output)
def test_starttag_junk_chars(self):
- self._parse_error("</>")
- self._parse_error("</$>")
+ self._run_check("</>", [])
+ self._run_check("</$>", [('comment', '$')])
self._parse_error("</")
self._parse_error("</a")
self._parse_error("<a<a>")
- self._parse_error("</a<a>")
+ self._run_check("</a<a>", [('endtag', 'a<a')])
self._parse_error("<!")
self._parse_error("<a")
self._parse_error("<a foo='bar'")
@@ -232,6 +232,44 @@
("endtag", "p"),
])
+ def test_invalid_end_tags(self):
+ # A collection of broken end tags. <br> is used as separator.
+ # see http://www.w3.org/TR/html5/tokenization.html#end-tag-open-state
+ # and #13993
+ html = ('<br></label</p><br></div end tmAd-leaderBoard><br></<h4><br>'
+ '</li class="unit"><br></li\r\n\t\t\t\t\t\t</ul><br></><br>')
+ expected = [('starttag', 'br', []),
+ # < is part of the name, / is discarded, p is an attribute
+ ('endtag', 'label<'),
+ ('starttag', 'br', []),
+ # text and attributes are discarded
+ ('endtag', 'div'),
+ ('starttag', 'br', []),
+ # comment because the first char after </ is not a-zA-Z
+ ('comment', '<h4'),
+ ('starttag', 'br', []),
+ # attributes are discarded
+ ('endtag', 'li'),
+ ('starttag', 'br', []),
+ # everything till ul (included) is discarded
+ ('endtag', 'li'),
+ ('starttag', 'br', []),
+ # </> is ignored
+ ('starttag', 'br', [])]
+ self._run_check(html, expected)
+
+ def test_broken_invalid_end_tag(self):
+ # This is technically wrong (the "> shouldn't be included in the 'data')
+ # but is probably not worth fixing it (in addition to all the cases of
+ # the previous test, it would require a full attribute parsing).
+ # see #13993
+ html = '<b>This</b attr=">"> confuses the parser'
+ expected = [('starttag', 'b', []),
+ ('data', 'This'),
+ ('endtag', 'b'),
+ ('data', '"> confuses the parser')]
+ self._run_check(html, expected)
+
def test_get_starttag_text(self):
s = """<foo:bar \n one="1"\ttwo=2 >"""
self._run_check_extra(s, [
diff --git a/Misc/NEWS b/Misc/NEWS
index 55b19b1..ce39863 100644
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -90,6 +90,8 @@
Library
-------
+- Issue #13993: HTMLParser is now able to handle broken end tags.
+
- Issue #13960: HTMLParser is now able to handle broken comments.
- Issue #9750: Fix sqlite3.Connection.iterdump on tables and fields