helix-query.yaml

version: 1
indices:
  articles:
    include:
      - /us/en/blog/**
      - /us/en/news/**
      - /us/en/library/**
      - /us/en/application/**
      - /us/en/info/**
    target: /us/en/article-index.json
    properties:
      lastModified:
        select: none
        value: parseTimestamp(headers["last-modified"], "ddd, DD MMM YYYY hh:mm:ss GMT")
      title:
        select: main h1
        value: textContent(el)
      navTitle:
        select: head > meta[name="navtitle"]
        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?:\/\/.*?(danaher.com))+(.*)|^(.*)")
      topics:
        select: head > meta[name="topics"]
        value: attribute(el, "content")
      readingTime:
        select: head > meta[name="readingtime"]
        value: attribute(el, "content")
      publishDate:
        select: head > meta[name="publishdate"]
        value: parseTimestamp(attribute(el, "content"), "ddd, DD MMM YYYY hh:mm:ss GMT")
      type:
        select: head > meta[name="template"]
        value: attribute(el, "content")
      brand:
        select: head > meta[name="brand"]
        value: attribute(el, "content")
  products:
    include:
      - /us/en/products/**
    exclude:
      - /us/en/products/family/**
      - /us/en/products/sku/**
    target: /us/en/products-index.json
    properties:
      title:
        select: main h1
        value: textContent(el)
      description:
        select: main .banner h2
        value: textContent(el)
      image:
        select: head > meta[property="og:image"]
        value: match(attribute(el, "content"), "^(https?:\/\/.*?(danaher.com))+(.*)|^(.*)")
      brand:
        select: head > meta[name="brand"]
        value: attribute(el, "content")
      type:
        select: head > meta[name="template"]
        value: attribute(el, "content")
      fullCategory:
        select: head > meta[name="fullcategory"]
        value: attribute(el, "content")
      lastModified:
        select: none
        value: parseTimestamp(headers["last-modified"], "ddd, DD MMM YYYY hh:mm:ss GMT")
  solutions:
    include:
      - /us/en/solutions/**
    target: /us/en/solutions-index.json
    properties:
      title:
        select: main .banner h1, main div:nth-child(4) h1
        value: textContent(el)
      solution:
        select: head > meta[name="solution"]
        value: attribute(el, "content")