- From: Stian Soiland-Reyes <soiland-reyes@cs.manchester.ac.uk>
- Date: Tue, 18 Nov 2014 16:02:31 +0000
- To: "David I. Lehn" <dil@lehn.org>
- Cc: Pemanent Identifier CG <public-perma-id@w3.org>
Had a quick go.. I seem to be onto something: https://travis-ci.org/stain/w3id.org/builds/41371490 https://github.com/stain/w3id.org/blob/travis-checker/.travis.yml Just perfecting the logging now before pull requesting. As you see the above caught a 404. Here's an example catching 500: https://travis-ci.org/stain/w3id.org/builds/41372440 It only tests the https://w3id.org/ URIs (but at http://localhost) that are in *.md (perhaps README.* instead ? ) - so to get your URL tested, simply add it to the README.md as several have done already (mainly myself..). I use a sed-like approach to get out just the URIs - I tried using the https://github.com/jch/html-pipeline to generate the HTML for the markdowns, but it basically requires 30 dependencies which truly is a bit overkill :) I also added a description of this to the top-level README.md. In theory Travis will automatically test and comment any github pull request. On 17 November 2014 17:23, David I. Lehn <dil@lehn.org> wrote: > On Mon, Nov 17, 2014 at 9:25 AM, Stian Soiland-Reyes > <soiland-reyes@cs.manchester.ac.uk> wrote: >> Do you think it is possible to have some kind of automatic testing or >> semi-automatic staging of a suggested .htaccess pull request? I'm kind >> of Travis-CI which can run tests on any pull request and then comment >> on the request with the test results. >> >> >> Perhaps it could be something as simple as >> >> cat $(find . -name README.md) | grep https://w3id | sed s/cleverbithere// >> >> and then try each of those to see that it gets redirected correctly >> (and that the destination is not 404)? >> > > If this worked, it would be nice, but it seems complex and a bit of > overkill for just testing redir rules. Adds overhead of having to add > test rules for what in most cases are simple redir rules. Tests could > also return 404 even though the rules are correct. Probably would > just want test system to test files related to new commits since old > rules may get stale and broken. Patches are of course welcome if you > have an idea how how to do this. :-) > > -dave -- Stian Soiland-Reyes, myGrid team School of Computer Science The University of Manchester http://soiland-reyes.com/stian/work/ http://orcid.org/0000-0001-9842-9718
Received on Tuesday, 18 November 2014 16:03:33 UTC