{#-
Theme override for the YazSes docs site.
Extends MkDocs Material's base to inject crawler directives and, if ever
needed, search-engine verification tags.
Google Search Console
---------------------
A URL-prefix property for https://mskazemi.com/ was verified on 2026-08-07 via
the DNS ("Domain name provider") method, which covers this /yazses/ subdirectory
and needs no meta tag at all. The token hooks below therefore stay EMPTY on
purpose — they exist only as a fallback if DNS verification is ever lost.
To use the fallback: add a URL-prefix property in GSC (NOT on
mskazemi.github.io — that host now 301-redirects here and serves no content),
pick the "HTML tag" method, and paste the token into mkdocs.yml under
extra.google_site_verification. Same pattern for Bing via
extra.bing_site_verification. Both are conditional, so an empty value emits
nothing and the build stays green.
-#}
{% extends "base.html" %}
{#- Only the `` tag. A design page's heading can be a whole sentence (a decision
record's is), so `hooks/design_tier.py` derives a short `seo_title` for the page's
front-matter dictionary. Setting `title` instead would also relabel Material's header
bar, which is visible UI; this block changes the search-result text and nothing else.
`home.html` overrides this block again for its own reason. -#}
{% block htmltitle %}
{% if page.meta and page.meta.seo_title %}
<title>{{ page.meta.seo_title }} - {{ config.site_name }}
{% else %}
{{ super() }}
{% endif %}
{% endblock %}
{% block extrahead %}
{{ super() }}
{#- Explicit crawler directives. Without this no robots meta is emitted at all:
indexing still defaults to allowed, but the snippet and image-preview
limits below are what let Google show a full-length snippet and a large
image rather than truncating both to its conservative defaults. -#}
{% if config.extra.google_site_verification %}
{% endif %}
{% if config.extra.bing_site_verification %}
{% endif %}
{#- Feed autodiscovery for the release notes (mkdocs-rss-plugin).
Without these two tags the feeds exist as files but nothing finds them:
browsers, feed readers and the crawlers that ingest feeds all look for
in the head, not for a guessable filename.
Absolute URLs via site_url because the site is served from the /yazses/
subdirectory of mskazemi.com — a root-relative "/feed_rss_created.xml"
would resolve to the wrong host path. -#}
{#- A compact, canonical product brief for answer engines that support the
llms.txt convention. Per-page Markdown alternates remain the primary
machine-readable surface; this link makes the site-level map discoverable. -#}
{% endblock %}