This link checker / sitemap generator is built using spatie/crawler which uses guzzle to do the crawling.

This is very similar to spatie/http-status-check but due to the way guzzle handles redirects I wasn't happy with the results; known 404's, and even whole chunks of known URLs were missing from crawl results. By following this https://github.com/guzzle/guzzle/blob/master/docs/faq.rst#how-can-i-track-redirected-requests I was able to get a crawl result with the complete list of responses as I was expecting.

The node.js test server is copied directly from spatie/http-status-check but refactored to offer a more diverse range of tests cases that cover the redirect issues noted above.

The examples below are run against the test server in this project.

EXCEPTION: Status code must be an integer value between 1xx and 5xx

I ran across this exception https://github.com/spatie/crawler/issues/271 and the patch for GuzzleHttp/Guzzle fixed it https://github.com/guzzle/guzzle/pull/2591 Without this patch CrawlerTest::testInvalidStatusCode will fail.

Follow redirects

By default spatie/http-status-check has guzzle set to not follow redirects. This results in the potential for parts of the site to be uncrawlable if they are behind a 301 or 302 redirect, and not linked internally anywhere else with a non-redirecting link.

Some webservers include a <a href="destination"> link on the 301/302 body and this will mitigate the problem (spaite follows and indexes), however if the webserver does not do this, then the link won't be followed to its true destination, and the destination won't be indexed.

This is most obvious with a redirect to a not found page: You'd expect to see a 404 here:

./http-status-check scan http://localhost:8080/redirectToNotFound

Start scanning http://localhost:8080/redirectToNotFound

[2020-02-22 12:07:22] 302 Found - http://localhost:8080/redirectToNotFound

Crawling summary
----------------
Crawled 1 url(s) with statuscode 302

Or a redirect to a page with links: You'd expect to see the rest of the site here:

$ ./http-status-check scan http://localhost:8080/redirectToFound

Start scanning http://localhost:8080/redirectToFound

[2020-02-22 12:08:34] 302 Found - http://localhost:8080/redirectToFound

Crawling summary
----------------
Crawled 1 url(s) with statuscode 302

If I enable RequestOptions::ALLOW_REDIRECTS as suggested here https://github.com/spatie/crawler/issues/263 I get:

$ ./http-status-check scan http://localhost:8080/redirectToNotFound

Start scanning http://localhost:8080/redirectToNotFound

[2020-02-22 12:13:53] 404: Not Found - http://localhost:8080/redirectToNotFound

Crawling summary
----------------
Crawled 1 url(s) with statuscode 404

and

$ ./http-status-check scan http://localhost:8080/redirectToFound

Start scanning http://localhost:8080/redirectToFound

[2020-02-23 18:10:19] 200 OK - http://localhost:8080/redirectToFound

Crawling summary
----------------
Crawled 1 url(s) with statuscode 200

This looks better, we can see the 404 and the 200 now, however on closer inspection you'll notice there are no 301 or 302's being shown. And the redirectToFound is actually showing as a 200 where the real response was 301, the 200 should be associated with the /found URL that is still missing from the results. For generating a sitemap of valid paged we'd need the destination URL, not the redirect URL.

Enter my sitemap crawler:

$ bob@chodbox:~/Projects/sitemap$ bin/crawler crawl http://localhost:8080/redirectToNotFound

302 http://localhost:8080/redirectToNotFound
404 http://localhost:8080/notFound

$ bin/crawler crawl http://localhost:8080/redirectToFound

302 http://localhost:8080/redirectToFound
200 http://localhost:8080/found

Collate foundOnUrls.

* Requires a patch to spatie/crawler https://patch-diff.githubusercontent.com/raw/spatie/crawler/pull/280 (auto applied) without this patch the `CrawlerTest::testInterlinked` test will fail.
* TODO: Document the confusing PART1&2 behaviour. (collating the redirect urls)