iartamonov/mautrix-telegram
1
0
Fork 0
Code Issues Pull Requests 1 Actions Packages Projects Releases Wiki Activity

Handle void tags correctly in the HTML parser. Fixes #309

Browse Source
This commit is contained in:
Tulir Asokan
2019-05-11 19:04:29 +03:00
parent 72a3cea948
commit 430f7b7217
1 changed files with 9 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
mautrix_telegram/formatter/from_matrix/html_reader.py
+9 -1
View File
@@ -29,6 +29,10 @@ class HTMLNode(list):
class NodeifyingParser(HTMLParser): class NodeifyingParser(HTMLParser):
# From https://www.w3.org/TR/html5/syntax.html#writing-html-documents-elements
void_tags = ("area", "base", "br", "col", "command", "embed", "hr", "img", "input", "link",
"meta", "param", "source", "track", "wbr")
def __init__(self): def __init__(self):
super().__init__() super().__init__()
self.stack = [HTMLNode("html", [])] # type: List[HTMLNode] self.stack = [HTMLNode("html", [])] # type: List[HTMLNode]
@@ -36,7 +40,11 @@ class NodeifyingParser(HTMLParser):
def handle_starttag(self, tag, attrs): def handle_starttag(self, tag, attrs):
node = HTMLNode(tag, attrs) node = HTMLNode(tag, attrs)
self.stack[-1].append(node) self.stack[-1].append(node)
self.stack.append(node) if tag not in self.void_tags:
self.stack.append(node)
def handle_startendtag(self, tag, attrs):
self.stack[-1].append(HTMLNode(tag, attrs))
def handle_endtag(self, tag): def handle_endtag(self, tag):
if tag == self.stack[-1].tag: if tag == self.stack[-1].tag: