- From: Michael[tm] Smith <mike@w3.org>
- Date: Wed, 10 Jul 2013 22:13:12 +0900
- To: Stéphane Corlosquet <scorlosquet@gmail.com>
- Cc: www-validator@w3.org
Stéphane Corlosquet <scorlosquet@gmail.com>, 2013-07-09 09:52 -0400: > Thanks Mike. Are there tests for the validator to avoid regressions? Yes, there are tests here: https://bitbucket.org/validator/syntax/src/default/relaxng/tests There's a test runner that gets called during the validator build, and that runs all the test through validation and reports (to stderr) any tests whose results are not what's expected. If you already have a built validator workspace, you can run the tests by evoking the validator build script like this: ./build/build.py test If you don't already have a validator build/test workspace, instructions for setting one up are here: https://bitbucket.org/validator/build/src > I would be happy to provide HTML documents featuring different kinds of > markup which are supposed to validate (and others which are not supposed > to validate). It'd be great to have a good set of tests for HTML+RDFa. So if you've got time to contribute some, I'd appreciate it. The best way to handle the test contributions would be for you to fork the validator "syntax" repo and then add your tests to that and send a pull request. https://bitbucket.org/validator/syntax/src Again, the actual tests themselves are here: https://bitbucket.org/validator/syntax/src/default/relaxng/tests Today I added "html5full-rdfalite" and "html5full-rdfa" directories there, with just a couple of tests so far in each. Any tests you add to the "html5full-rdfalite" directory will be checked by the test runner against RDFa Lite 1.1 conformance rules, and any tests you add to the "html5full-rdfa" directory will be checked against full RDFa 1.1 conformance rules. Within those directories there are subdirectories named "valid/" and "invalid/". For a test in the "valid/" subdirectory, if the test runner gets an error for it during validation, it'll just report the error as-is to stderr. For a test in the "invalid/" subdirectory, if the test runner doesn't get an error for it during validation, the test runner will report "file foo was supposed to be invalid but was not" to stderr. Test files with an .html extension are parsed as text/html, and test files with an .xhtml extensions are parsed as XML. That's about it. If you have any questions, just let me know. --Mike -- Michael[tm] Smith http://people.w3.org/mike
Received on Wednesday, 10 July 2013 13:13:27 UTC