Open petdance opened 6 years ago
@petdance yes, I think this is a bug, and needs to be fixed
This seems yet another case where tidy
needs to support the current HTML5 spec...
I can read in style docs <style>
can be used In the body, where flow content is expected.
... although there are some warning type notes about it...
One quick patch I experimented with was -
diff --git a/src/clean.c b/src/clean.c
index e96dd3f..8023616 100644
--- a/src/clean.c
+++ b/src/clean.c
@@ -2747,7 +2747,7 @@ static void StyleToHead(TidyDocImpl* doc, Node *head, Node *node, Bool fix, int
TY_(InsertNodeAtEnd)(head, node); /* add to end of head */
TY_(Report)(doc, node, head, MOVED_STYLE_TO_HEAD); /* report move */
}
- else
+ else if (!TY_(IsHTML5Mode)(doc)) /* Is. #730 - style allowed in body */
{
TY_(Report)(doc, node, head, FOUND_STYLE_IN_BODY);
}
But maybe this is not the complete story... because this would allow it to be the child
of anything... maybe there needs to be more checking... although, with --fix-style-tags no
it does fix the sample given...
Note such a fix may also mean adjusting, changing the fix-style-tags
documentation, or its yes
default, and/or the MOVED_STYLE_TO_HEAD
and the FOUND_STYLE_IN_BODY
messages...
Look forward to feedback, patches, or PR... thanks...
I don't have any feedback one way or the other. I would just expect that if you turn on one flag, it wouldn't suggest you turn on a different one.
Here's an HTML file with
<style>
tag outside of the<head>
.Tidy complains about it.
So I do what tidy says and it complains a different complaint.
What can I do to make tidy happy about this situation?