internal error when displaying .sql attachment

[anonymous]

Trac returns an internal error when rendering a .sql attachment containing the following SQL statement:

INSERT INTO `somedb` VALUES ('<a "http://trac.edgewall.org">trac</a>');

I think this is valid SQL?

The error:

Traceback (most recent call last):
  File "/usr/lib/python2.4/site-packages/trac/web/main.py", line 387, in dispatch_request
    dispatcher.dispatch(req)
  File "/usr/lib/python2.4/site-packages/trac/web/main.py", line 244, in dispatch
    req.display(template, content_type or 'text/html')
  File "/usr/lib/python2.4/site-packages/trac/web/api.py", line 358, in display
    data = self.hdf.render(template, form_token)
  File "/usr/lib/python2.4/site-packages/trac/web/clearsilver.py", line 296, in render
    injector.feed(template.render())
  File "/usr/lib/python2.4/HTMLParser.py", line 108, in feed
    self.goahead(0)
  File "/usr/lib/python2.4/HTMLParser.py", line 150, in goahead
    k = self.parse_endtag(i)
  File "/usr/lib/python2.4/HTMLParser.py", line 314, in parse_endtag
    self.error("bad end tag: %r" % (rawdata[i:j],))
  File "/usr/lib/python2.4/HTMLParser.py", line 115, in error
    raise HTMLParseError(message, self.getpos())
HTMLParseError: bad end tag: '</a">', at line 55, column 282

[anonymous]

The above link to the attachment page returns the same error.

The attachment may be downloaded directly here:

test.sql

[anonymous]

the following can be attached as both test.txt and test.sql:

'<a href="https://"></a>'

the following can only be attached as test.txt:

'<a href="http://"></a>'

changing https to http prevents rendering as test.sql.

[cboos]

Well, this is a known problem with 0.10.1 onwards, a side-effect of #4931.

Broken HTML produced by (broken) syntax highlighters can't be parsed and will trigger such an error. This is similar to the problem already reported in #4365.

[anonymous]

thanks. i should have searched more carefully.

ticket:4931 is related to SMTP and TLS?

[cboos]

Oops, got the reference totally wrong, sorry: I meant a side-effect of #4049 (#4931 has nothing to do with the issue).