Extract the metadata title from a header

[4l1fe]

Hey. Most of the time i name files shortly and write the fuul title inside a file content and would like a page have slug and title extracted from it, it's like <h1> in html

Is it possible to set up?

[plgruener]

What format do you write your pages in? If markdown/rst, you should rather just use the :title: metadata field (inside each file) and adjust your template to automatically use that as the page title/ format it as <h1>{{page.title}}</h1> or something. You may also need to turn off the 'get title from filename' option.
If you write directly in html, I think there's also option to embed metadata in there (via <meta> tags, but not sure).

[mcepl]

Well, except with rST this works:

Harry and Ginny as siblings
###########################

:date: 2019-07-15T23:07:27
:category: literature
:tags: storyPrompt, harryPotter, plotBunny, fromReddit

Originally published as “`Harry & Ginny are brother and sister`_”
on Reddit.

and the post has both slug, and metadata title generated from that rST headline.
https://matej.ceplovi.cz/blog/harry-and-ginny-as-siblings.html

Except if I try to do the same with Markdown:

# Harry and Ginny as siblings

Date: 2019-07-15T23:07:27
Category: literature
Tags: storyPrompt, harryPotter, plotBunny, fromReddit

Originally published as "[Harry & Ginny are brother and
sister](https://www.reddit.com/r/HPfanfiction/comments/bevfuu/lf_harry_ginny_are_brother_and_sister/)"
on Reddit.

I get an error:

/archiv/knihovna/blog/publishconf.py -t "/home/matej/repos/pelican-themes/zurb-F5-basic"
[00:11:44] ERROR    Skipping                                                               contents.py:176
                    /home/matej/archiv/knihovna/blog/content/literature/hinny_siblings.md:                
                    could not find information about 'title'

[justinmayer]

I do not use reStructuredText enough to understand why it supports this use case, but the code makes it clear that every page must have a title, and the aforementioned error is raised when mandatory metadata is missing.

I think perhaps a HEADING_METADATA setting could be a way to add support for this use case, where if Title metadata is not explicitly defined, maybe setting HEADING_METADATA = 0 uses the first heading as the title, a value of -1 uses the last heading as the title, and HEADING_METADATA = None (default) would keep the current behavior. Just an idea.

Setting aside the implementation details, I think this might be a good candidate as a Pelican plugin, for whomever would like to add support for this use case.

[mcepl]

See #3539 what I have vibe-coded on the topic.