Strip tags from page title

[waylan]

There shouldn't be any HTML tags in an HTML <title> so there is no harm in stripping them. And we get the benefit of not having the tags rendered in the browser's title bar/tab. See this comment in Python-Markdown/markdown#1379.

[oprypin]

Something I noticed - probably in <link rel="next" it should get the same treatment

mkdocs-nature/nature/base.html

Lines 149 to 150 in 5b785b5

	{% if page and page.next_page %}<link rel="next" title="{{ page.next_page.title }}" href="{{ page.next_page.url|url }}" />{% endif %}
	{% if page and page.previous_page %}<link rel="prev" title="{{ page.previous_page.title }}" href="{{ page.previous_page.url|url }}" />{% endif %}

Anyway this is partly a new issue with MkDocs 1.5 and I'm looking into it generally, maybe this shouldn't be necessary mkdocs/mkdocs#3357 (comment)

[waylan]

Actually in our case we don't want to strip the <code> tags on the page. Only in the <title> tag where HTML does not get rendered does this matter. So actually, the raw HTML not being stripped is desirable in our case. In fact, we added the <code> tags specifically to avoid tripping up the spellchecker in our CI tests (as the contents of <code> tags are skipped). Therefore, if the current behavior was reverted, we would lose a feature.

I realize that you may have inadvertently introduced a breaking change here, but I'm inclined to suggest that you leave it as-is and document that themes should use the striptags filter where needed. But then again, it may be that that doesn't work for all edge cases.

[oprypin]

@waylan Thanks for your input, you can comment on the issue too.

But what I'm saying is still true and separate-

title="{{ page.next_page.title }}" <- raw HTML gets thrown into the quoted part here (and Firefox at least underlines it in red in view-source:), it should in fact be stripped same as inside <title>

In the body sure, you don't need to strip it

[waylan]

Ah, right, in the title. Yeah, that is an issue. Thanks for pointing that out.