Add :gitauthordate and :gitcommitdate front matter config tokens

[dilyanpalauzov]

The generated sitemap.xml includes for each post a tag <lastmod>DATE-TIME</lastmod>. The included date should be the last time, when the page was modified. With enableGitInfo: true the last modification is derived from Author Date: of the commit. At https://gohugo.io/methods/page/gitinfo/ is written:

When you set enableGitInfo to true, or enable the feature with the command line flag, the last modification date for each content page will be the Author Date of the last commit for that file.

This is configurable. See https://gohugo.io/configuration/front-matter/#dates .

My reading of "This is configurable" means that I can use CommitDate instead of AuthorDate. But I cannot out of the box do so.

When I modify a commit with an article several times, the AuthorDate of the commit does not change, the CommitDate changes. The implication is that with enableGitInfo the generated site may contains under <lastmod> wrong information: AuthorDate instead of CommitDate and thus not the last modification.

• Provided that CommitDate cannot be before AuthorDate, when generating sitemaps with enableGitInfo: true the date in <lastmod> must be from CommitDate.

Disadvantage of this suggestion: with git-rebase CommitDate changes, while the article and AuthorDate does not change. However this is insigificant: too-new last-modification in the sitemap will cause recrawling the web page by search engines. While too-old last-modification in the sitemap file will skip recrawling of updated content.

hugo v0.148.0-DEV-1e9a0b93e3e5e1e0d25d4b8d8c36d5a2d7656f66+extended linux/amd64 BuildDate=2025-07-03T08:46:37Z

[jmooring]

I'd argue that a material content change (one that warrants a new crawl) should be a new commit, not an amendment.

To use the commit date in sitemaps, override the embedded sitemap template with something like this:

layouts/sitemap.xml

{{ printf "<?xml version=\"1.0\" encoding=\"utf-8\" standalone=\"yes\"?>" | safeHTML }}
<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9"
  xmlns:xhtml="http://www.w3.org/1999/xhtml">
  {{ range where .Pages "Sitemap.Disable" "ne" true }}
    {{- if .Permalink -}}
  <url>
  
    {{ $lastmod := .Lastmod }}
    {{ with .GitInfo }}
      {{ $lastmod = .CommitDate }}
    {{ end }}

    <loc>{{ .Permalink }}</loc>{{ if not $lastmod.IsZero }}
    <lastmod>{{ safeHTML ( $lastmod.Format "2006-01-02T15:04:05-07:00" ) }}</lastmod>{{ end }}{{ with .Sitemap.ChangeFreq }}
    <changefreq>{{ . }}</changefreq>{{ end }}{{ if ge .Sitemap.Priority 0.0 }}
    <priority>{{ .Sitemap.Priority }}</priority>{{ end }}{{ if .IsTranslated }}{{ range .Translations }}
    <xhtml:link
                rel="alternate"
                hreflang="{{ .Language.LanguageCode }}"
                href="{{ .Permalink }}"
                />{{ end }}
    <xhtml:link
                rel="alternate"
                hreflang="{{ .Language.LanguageCode }}"
                href="{{ .Permalink }}"
                />{{ end }}
  </url>
    {{- end -}}
  {{ end }}
</urlset>

As you noted above, the date that content was "last modified" is sometimes the author date and sometimes the commit date. To keep things simple we picked one (author date), which in most cases will be accurate.

I suppose we could add a couple of front matter tokens:

• :gitauthordate (:git would be an alias to this)
• :gitcommitdate

Then site authors can configure this site wide, but we'd keep the existing default.

[bep]

When I modify a commit with an article several times, the AuthorDate of the commit does not change, the CommitDate changes. The implication is that with enableGitInfo the generated site may contains under wrong information: AuthorDate instead of CommitDate and thus not the last modification.

In my head, semantically, Git's author date is what comes closest to lastMod. That is also what GitHub uses in its listings. If you want to, you can get that up to date with:

git commit --amend --date=now --no-edit