Use a different approach to fix the smarty bug #419
Conversation
That commit caused a regression where `<<` and `>>` inside code blocks were wrongly replaced with double quotes. This reverts commit 5029d82.
markdown/inlinepatterns.py
Outdated
There was a problem hiding this comment.
I think this is a sensible approach when the smarty extension is in use, but the default behavior should not be changed when smarty is not enabled. I would expect that the smarty extension would replace the HTML pattern with a new one if angle quotes are enabled. Otherwise, leave it unchanged.
And we should have a test for each state (enabled and not enabled). The not enabled state should return the same results as we would get without smarty before this patch was applied.
There was a problem hiding this comment.
Is <foo>> really a valid HTML tag?
There was a problem hiding this comment.
No, of course not. But the parser is not concerned with whether raw HTML is "valid". Without smarty's angle quotes enabled, I expect the same default behavuor we have now: <foo>> becomes <foo>>.
|
@waylan I have now updated the pull request as you suggested. |
Add a restriction that the closing angled quote should not be duplicate, so that we can use our own pattern for handling duplicate quotes.
Use a different approach to fix the smarty bug
Just update
HTML_REso that it doesn't act on<<...>>.