This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
urltestdata.txt has the following tests that don't match the production rules for valid URIs but don't produce any conformance errors. This can be fixed with two changes: A parse error if leading/trailing white space is present in: https://url.spec.whatwg.org/#concept-basic-url-parser A parse error if a colon is present in: https://url.spec.whatwg.org/#relative-path-start-state - - - The corresponding places where the rewritten parser would have to change would be: https://specs.webplatform.org/url/webspecs/develop/#concept-basic-url-parser https://specs.webplatform.org/url/webspecs/develop/#path
Here are the tests from urltestdata.txt: "\t :foo.com \n" " foo.com " " \t" ":foo.com/" ":" ":a" ":/" ":#" ":23" "::" "::23" Based on discussions with annevk, this should be fixed in https://specs.webplatform.org/url/webspecs/develop/#relative-url instead; in the description for row #4, add a check for the first segment containing a colon and indicating a conformance error if one is found.
Fixed by https://github.com/webspecs/url/commit/a6114157782a7f3d260e2a8d2e8db2e914633f3f
Fix was only applied to webspecs
Thank you. It's not entirely clear to me why a leading colon is problematic, but I've fixed the other missing parse errors and some. If you find out why a leading colon is problematic a GitHub issue would be appreciated. https://github.com/whatwg/url/commit/0bae5e5608bfc6132937f8f28388c6f796c9a846