Check Website Dependencies

This action uses the website oracle to check the dependencies of a website. It will fail if any of the dependencies are broken so that you can detect them before end users are affected by them. This action currently supports the following dependency types.

• page - Links to other pages which user can navigate to
• script - Links to script files
• stylesheet - Links to stylesheets

Website Oracle can start at a page of your choice and scrape through your website, moving from one page to another, to ensure that links are not broken.

Inputs

Input	Description	Default Value
apiKey	API Key for Website Oracle	N/A (Required)
url	The URL where the dependency analysis should start	N/A (Required)
maxDepth	The maximum number of times the dependencies should be fetched recursively (The maximum depth of the dependency tree)	1
includeTypes	The types of dependencies to analyze as a comma seperated list. It should be one of (page, script, stylesheet)	All types
config	The path to the config JSON file, relative to the repository root	.github/actions/website-oracle.json

Website Oracle is currently available through Rapid API. Subscribe to the API and get a token from Rapid API which can be used to call the Website Oracle.

Configuration File

The configuration file should contain a JSON object with the fields.

Field	Type	Default Value
scrapeRules	Array of Scrape Rule objects	[]

Example:

{
  "scrapeRules": [
    {
      "urlMatch": {
        "value": " https://www.googletagmanager.com/gtag/js?id=.*",
        "type": "regex"
      },
      "action": "ignore"
    }
  ]
}

Scrape Rule Config

The scrape rule configuration object should contain JSON objects with the following fields.

Field	Description	Default Value
urlMatch	URL Match object	N/A (Required)
action	One of ignore, checkStatusOnly scrape	ignore

The action values should be specified based on what you wish to do when a URL is matched by this rule.

• ignore - Ignore the URL completely without even checking the status.
• checkStatusOnly - Check the status of the URL, but do not get the dependencies of the URL itself.
• scrape - Check the status of the URL and then continue scraping of the dependencies of the URL (subject to maxDepth and includeTypes).

URL Match Config

The URL match configuration object should contain JSON objects with the following fields.

Field	Type	Default Value
value	The exact URL or the Regex depending on the URL match type	N/A (Required)
type	One of exact, regex	exact

The URL match type should be specified based on what you wish to do when a URL is tested against a rule.

• exact - The value should match exactly to the URL
• regex - The value is used as a regular expression to test the URL for a string match

Examples

Please us the <VERSION> with the exact version tag you wish to use.

name: Check Dependencies

on:
  workflow_run:
    workflows:
      - pages-build-deployment
    types:
      - completed
  schedule:
    - cron: 0 0 * * SUN

jobs:
  check-dependencies:
    runs-on: ubuntu-latest
    name: Check Dependencies
    steps:
      - name: Checkout Repository
        uses: actions/checkout@v3
      - name: Check Dependencies
        uses: cadmus-labs/check-website-dependencies-action@<VERSION>
        with:
          apiKey: ${{ secrets.WEBSITE_ORACLE_API_KEY }}
          url: "https://cadmus-labs.github.io/"