We already do this in a few cases, e.g. tests/phpunit/Parsoid/Fixer/WikiLinkFixerTest.php. See also T76586: K3. Run Flow tests with Parsoid on Jenkins.

A bunch of that work was started in https://gerrit.wikimedia.org/r/#/c/191697/, but we abandoned it.

I still don't think we should mock Parsoid response. We shouldn't pretend Parsoid is a service we can just swap out, and the things we do with Parsoid's response (output/store it, extract some nodes) are very simple. There's little logic on our end, other than interpreting the data in the way Parsoid outputs it. The most likely source of regression is when Parsoid changes its output. If we mock those responses, we wouldn't be aware of any such changes.

And we can already run tests without Parsoid - we don't *really* require it to be present (which unfortunately means we have to support 2 parsers)

Personally, I think it's more worthwile to spend time letting us connect to Parsoid from Jenkins, than mocking up Parsoid responses.

In T103041#1381453, @matthiasmullie wrote:

Personally, I think it's more worthwile to spend time letting us connect to Parsoid from Jenkins, than mocking up Parsoid responses.

Agreed.

Hi I dont know if this is the riight pace to ask. Some people on the francophone wikipedia are receiving error message on flow discussion pages
"Erreur d’accès au serveur Parsoid/RESTBase (HTTP 400"
This issue is also duscussud on discord by confirmed contributors. What can be done?

@Nattes: Please follow https://www.mediawiki.org/wiki/How_to_report_a_bug and create a separate task, as that's a different issue. Thanks!

@Nattes: Maybe T249997: Flow failing with Error contacting the Parsoid/RESTBase server (HTTP 400) ?