Closing this as a duplicate of #25. It is not fixed, and I want to be clear about that rather than have the closure read as a resolution.

@technopagan called it correctly in 2020: this and #25 are the same defect. graphtage/xml.py:12 imports xml.etree.ElementTree, and HTML(XML) at xml.py:443 overrides nothing but __init__, so HTML goes through the XML parser unchanged. HTML void elements are never closed, so the parser reaches a closing tag with an unclosed <meta> or <br> still on its stack and blames the closing tag:

$ cat t.html
<html><head><title>t</title></head><body><br></body></html>
$ graphtage --no-status t.html t.html
Error parsing t.html: mismatched tag: line 1, column 47

Column 47 is </body>, which is well formed. The real problem is the <br> before it. Same root cause as the unquoted attribute in #25, and one parser swap fixes both.

PR #106 is the tracked fix. It is open with changes requested and currently conflicts with master, so it needs a rebase before it can go anywhere. I am consolidating on #25 so there is one place to follow.

Uh oh!

Fails to match closing HTML tags #26

Description

Activity

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions