Skip to main content
Version: v2

Cross-Site Search

Search for a device across other measurement databases from your own site's search box. Results link straight to the matching site with the device loaded.

Overview​

When a visitor searches for a device you don't have, modernGraphTool can show them who does. Below your own results, it lists matches from other public measurement databases, grouped by site and rig type.

This works on any host — squig.link is not required. Cross-site search reads a single prebuilt index document, so it costs one request no matter how many databases are listed.

The index comes from GraphAggregator, a community project that crawls known measurement databases and publishes their device lists as one JSON file. modernGraphTool ships with its official URLs built in, so there is nothing to set up.

Configuration​

Configured in config.js:

CROSS_SITE_SEARCH: {
    ENABLED: true,
    INDEX_URLS: [],
    SQUIGLINK_FALLBACK: true,
}

Configuration Options​

  • ENABLED: Toggle cross-site search. Defaults to on.
  • INDEX_URLS: Index documents to try, in order. Leave empty to use the official GraphAggregator index (with its GitHub Pages mirror as an automatic backup).
  • SQUIGLINK_FALLBACK: When no index is reachable and the site runs on squig.link, fall back to crawling each squig.link site's phone_book.json. This is the legacy path: it works, but it costs one request per database.

Self-hosting the index​

If you'd rather not depend on a third-party origin, publish your own index document and point INDEX_URLS at it:

CROSS_SITE_SEARCH: {
    ENABLED: true,
    INDEX_URLS: [
        'https://example.com/my-index.json',
        'https://backup.example.com/my-index.json'
    ],
}

Each URL is tried in order and the first one that returns a valid document wins, so a mirror can be listed as a backup. The document must follow the GraphAggregator schema — see the GraphAggregator repository for the spec and for tooling that generates it.

Notes​

  • Databases hosted on the site the visitor is already on are filtered out of the results.
  • Results are capped at 100 per query. When more match, a hint invites the visitor to refine the search.
  • The index is fetched once, lazily, the first time a visitor types at least two characters — it costs nothing for visitors who never search.