Support pageExtensions in no-html-link-for-pages eslint rule

[boris324]

Fixes #53473

The no-html-link-for-pages ESLint rule currently has hardcoded file extension matching (tsx, ts, jsx, js) in the URL parsing functions. This means projects using custom pageExtensions (like mdx) in their Next.js config have those pages silently ignored by the lint rule.

This PR makes the internal parseUrlForPages and parseUrlForAppDir functions accept an optional pageExtensions parameter, building the file extension regex dynamically instead of using the hardcoded pattern. The rule reads pageExtensions from context.settings.next (same pattern used by getRootDirs for rootDir) and passes it through.

Users can configure it in their ESLint config:

{
  "settings": {
    "next": {
      "pageExtensions": ["tsx", "ts", "jsx", "js", "mdx"]
    }
  }
}

When not configured, defaults to the original ["tsx", "ts", "jsx", "js"] behavior.

Added tests covering custom page extensions detection and the default fallback behavior.

[nextjs-bot]

Allow CI Workflow Run

• approve CI run for commit: 7db932d

Note: this should only be enabled once the PR is ready to go and can only be enabled by a maintainer