helix-query.yaml

version: 1

indices:
  site-en: &base-site
    include:
      - '/ext/en/**'
    exclude:
      - '/ext/en/drafts'
      - '/ext/en/drafts/**'
      - '/ext/en/fragments'
      - '/ext/en/fragments/**'
      - '/ext/en/footer'
      - '/ext/en/nav'
      - '/ext/en/data'
      - '/ext/en/data/**'
    target: /ext/en/query-index.json
    properties:
      title:
        select: main h1
        value: textContent(el)
      navTitle:
        select: head > meta[property="og:title"]
        value: attribute(el, "content")
      description:
        select: head > meta[name="description"]
        value: attribute(el, "content")
      image:
        select: head > meta[property="og:image"]
        value: match(attribute(el, "content"), "^(https?:\/\/.*?(worldbank.org))+(.*)|^(.*)")
      publishDate:
        select: head > meta[name="published-time"]
        value: parseTimestamp(attribute(el, "content"), "ddd, DD MMM YYYY hh:mm:ss GMT")
      contentType:
        select: head > meta[name="content_type"]
        value: attribute(el, "content")
      languages:
        select: head > meta[name="languages"]
        value: attribute(el, "content")
      originating_unit:
        select: head > meta[name="originating_unit"]
        value: attribute(el, "content")
      author:
        select: head > meta[name="author"]
        value: attribute(el, "content")
      information_classification:
        select: head > meta[name="information_classification"]
        value: attribute(el, "content")
      lastModified:
        select: none
        value: parseTimestamp(headers["last-modified"], "ddd, DD MMM YYYY hh:mm:ss GMT")

  site-zh:
    <<: *base-site
    include:
      - '/ext/zh/**'
    exclude:
      - '/ext/zh/drafts'
      - '/ext/zh/drafts/**'
      - '/ext/zh/fragments'
      - '/ext/zh/fragments/**'
      - '/ext/zh/footer'
      - '/ext/zh/nav'
      - '/ext/zh/data'
      - '/ext/zh/data/**'
    target: /ext/zh/query-index.json
  
  site-ru:
    <<: *base-site
    include:
      - '/ext/ru/**'
    exclude:
      - '/ext/ru/drafts'
      - '/ext/ru/drafts/**'
      - '/ext/ru/fragments'
      - '/ext/ru/fragments/**'
      - '/ext/ru/footer'
      - '/ext/ru/nav'
      - '/ext/ru/data'
      - '/ext/ru/data/**'
    target: /ext/ru/query-index.json

  site-fr:
    <<: *base-site
    include:
      - '/ext/fr/**'
    exclude:
      - '/ext/fr/drafts'
      - '/ext/fr/drafts/**'
      - '/ext/fr/fragments'
      - '/ext/fr/fragments/**'
      - '/ext/fr/footer'
      - '/ext/fr/nav'
      - '/ext/fr/data'
      - '/ext/fr/data/**'
    target: /ext/fr/query-index.json

  site-es:
    <<: *base-site
    include:
      - '/ext/es/**'
    exclude:
      - '/ext/es/drafts'
      - '/ext/es/drafts/**'
      - '/ext/es/fragments'
      - '/ext/es/fragments/**'
      - '/ext/es/footer'
      - '/ext/es/nav'
      - '/ext/es/data'
      - '/ext/es/data/**'
    target: /ext/es/query-index.json

  site-ar:
    <<: *base-site
    include:
      - '/ext/ar/**'
    exclude:
      - '/ext/ar/drafts'
      - '/ext/ar/drafts/**'
      - '/ext/ar/fragments'
      - '/ext/ar/fragments/**'
      - '/ext/ar/footer'
      - '/ext/ar/nav'
      - '/ext/ar/data'
      - '/ext/ar/data/**'
    target: /ext/ar/query-index.json