You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently we use a parser written in JS, using ohm.js, to parse the wpt.fyi search syntax.
However, this ties the search syntax to the web front end, and there's many cases where it would be useful to be able to take a wpt.fyi URL from a user and then perform the query again.
Currently to do this requires extracting a bunch of code from webapp/components/test-search.js and then running the JS locally. Even just refactoring this so that TestSearch.parseAndInterpretQuery is something that can be used without including Polymer etc. would be a win, so that such a file can be more directly copied.
Ideally we'd move the parsing to the server-side, so other clients can be given wpt.fyi URLs, extract the query parameter, and then make the request themselves. It's unclear what we gain from the parser being client-side?
The text was updated successfully, but these errors were encountered:
I don't see much value in the current setup and agree that we would be better served with a backend query parser, similar to what we use in other projects.
Currently we use a parser written in JS, using ohm.js, to parse the wpt.fyi search syntax.
However, this ties the search syntax to the web front end, and there's many cases where it would be useful to be able to take a wpt.fyi URL from a user and then perform the query again.
Currently to do this requires extracting a bunch of code from
webapp/components/test-search.js
and then running the JS locally. Even just refactoring this so thatTestSearch.parseAndInterpretQuery
is something that can be used without including Polymer etc. would be a win, so that such a file can be more directly copied.Ideally we'd move the parsing to the server-side, so other clients can be given wpt.fyi URLs, extract the query parameter, and then make the request themselves. It's unclear what we gain from the parser being client-side?
The text was updated successfully, but these errors were encountered: