Deal more appropriately with bare ampersands and pointy brackets; this
module has to deal with "class" HTML-as-deployed as well as XHTML, so we
cannot be as strict as XHTML allows.
This closes SF bug #453059, but uses a different fix than suggested in
the bug comments.
diff --git a/Lib/HTMLParser.py b/Lib/HTMLParser.py
index 39a5d82..954ce26 100644
--- a/Lib/HTMLParser.py
+++ b/Lib/HTMLParser.py
@@ -15,7 +15,8 @@
interesting_normal = re.compile('[&<]')
interesting_cdata = re.compile(r'<(/|\Z)')
-incomplete = re.compile('&([a-zA-Z][-.a-zA-Z0-9]*|#[0-9]*)?')
+incomplete = re.compile('&([a-zA-Z][-.a-zA-Z0-9]*'
+ '|#([0-9]*|[xX][0-9a-fA-F]*))?')
entityref = re.compile('&([a-zA-Z][-.a-zA-Z0-9]*)[^a-zA-Z0-9]')
charref = re.compile('&#(?:[0-9]+|[xX][0-9a-fA-F]+)[^0-9a-fA-F]')
@@ -185,11 +186,8 @@
elif declopen.match(rawdata, i): # <!
k = self.parse_declaration(i)
else:
- if i < n-1:
- raise HTMLParseError(
- "invalid '<' construct: %s" % `rawdata[i:i+2]`,
- self.getpos())
- k = -1
+ self.handle_data("<")
+ k = i + 1
if k < 0:
if end:
raise HTMLParseError("EOF in middle of construct",
@@ -203,7 +201,7 @@
self.handle_charref(name)
k = match.end()
if rawdata[k-1] != ';':
- k = k-1
+ k = k - 1
i = self.updatepos(i, k)
continue
match = entityref.match(rawdata, i)
@@ -212,17 +210,19 @@
self.handle_entityref(name)
k = match.end()
if rawdata[k-1] != ';':
- k = k-1
+ k = k - 1
i = self.updatepos(i, k)
continue
- if incomplete.match(rawdata, i):
- if end:
+ match = incomplete.match(rawdata, i)
+ if match:
+ rest = rawdata[i:]
+ if end and rest != "&" and match.group() == rest:
raise HTMLParseError(
"EOF in middle of entity or char ref",
self.getpos())
return -1 # incomplete
- raise HTMLParseError("'&' not part of entity or char ref",
- self.getpos())
+ self.handle_data("&")
+ i = self.updatepos(i, i + 1)
else:
assert 0, "interesting.search() lied"
# end while