Search engines and link previews

Every page of a site tells search engines and link previews what it is. The portal writes all of this from the site’s files, so a page needs nothing more than its title to be found. A site can say more.

What each page says

• Its description: the page’s own, else the site’s.
• Its address: one address per language. The home page’s address is the site’s root, such as https://example.ro/ and https://example.ro/en.
• The same page in the site’s other languages, so search engines show each visitor theirs.
• How it looks when shared: its title, the site’s name, its description and its image.
• Structured data about the site and the page, and about the organization behind the site when it names one.

Describe a page

In the site’s files, a page’s seo holds:

yaml

seo:
  description: { ro: Programul și serviciile primăriei., en: The city hall's hours and services. }
  image: https://img.example.ro/share.png
  noindex: false

• description: at most 300 characters. Once it is given in one language, give it in all of the site’s languages.
• image: an https: address. It shows when the page is shared.
• noindex: true: asks search engines not to list the page.

Administrators can set these in the console too, under Search and sharing in the page editor.

Describe a site

A site’s site.yaml can say the same for all its pages, and who is behind it:

yaml

seo:
  description: { ro: Site-ul Primăriei Exemplu., en: Example City Hall's site. }
  image: https://img.example.ro/share.png
  organization:
    name: Primăria Exemplu
    url: https://example.ro
    logo: https://img.example.ro/logo.png
  index: true

A page’s own description and image win over the site’s. index: false asks search engines to leave the whole site out.

What crawlers are told

Each address the portal serves answers two files:

• /robots.txt keeps crawlers away from the portal’s own paths, but not from what pages need to show, and names the sitemap.
• /sitemap.xml lists every page anyone may see, in each language, with its other-language addresses. It leaves out:
  • pages only signed-in people may see;
  • pages with noindex;
  • the pages of a site with index: false.

Previews are never listed: their robots.txt keeps every crawler out.