Wrong URL in admin pagination when using CloudFlare Workers for Subdirectory

Question

We implemented a cloudflare worker to serve our subdomain blog blog.domain.com via a subdirectory www.domain.com/blog it's done as described here https://blog.cloudflare.com/subdomains-vs-subdirectories-improved-seo-part-2/

We updated the site settings in WP and all works fine but the pagination in admin is still pointing to the original host which is blog.domain.com

I see similar issues and questions being posted when using a proxy to serve the blog from a special directory, it seems to be no real solution other than modifying the WP core files which I would avoid.

Where is WP pulling the host other than the wp-config.php file?

Answer 1

With a modified version of the CF Workers, the issue with pagination has been resolved. The idea I found in the comments of the cloudflare post which points to this workers script

    addEventListener('fetch', event => {
  event.respondWith(handleRequest(event.request))
})

// keep track of all our blog endpoints here
const config = {
  hostname: "blog.domain.com",
  mainhostname: "www.domain.com",
  targetSubdirectory: "/blog",
  assetsPathnames: ["/wp-content/", "/wp-includes/"]
}

async function handleRequest(request) {
  const parsedUrl = new URL(request.url)
  const requestMatches = match => new RegExp(match).test(parsedUrl.pathname)
  
    /**
     * Best practice is to only assign new properties on the request
     * object (i.e. RequestInit props) through either a method or the constructor
     */
    let newRequestInit = {
    }
    // Change URL
    let url = parsedUrl.href.replace(parsedUrl.hostname + config.targetSubdirectory, config.hostname)
    if (url.includes('/2020/06/01/some-post') && url[url.length-1] !== '/') {
      url = url + '/'
    }
    // Change just the host
    url = new URL(url)
    // Best practice is to always use the original request to construct the new request
    // thereby cloning all the attributes, applying the URL also requires a constructor
    // since once a Request has been constructed, its URL is immutable.
    const newRequest = new Request(url, new Request(request, newRequestInit))
    try {
      return await fetch(newRequest)
    } catch (e) {
      return new Response(JSON.stringify({ error: e.message }), { status: 500 })
    }
  }

p.s. there is one line to fix URLs without a slash which did not work with the out of the box script—but we should be able to modify the js code to handle such situations without hardcoding a single URL