Martin probably will not have time to work on this for 7.4, reassigning to Petr. We should try to do this for 7.4 - changing to DEFECT.

I think that it's needed to create map of every property and method of the JS documentation or at least of the base objects. There is a problem that every part of the documentation is taken from another source (there is no single good source which would gather all these docs).

It means that there must be introduced method which would guess the link (including i.e. URI subpath or #anchor) to the specific field according of the base JS element type (Core JS, DOM JS). But it's possible that we will not be able to implement it generally for objects and mapping for every field/method will be necessary.

Thanks for the evaluation Martin. Of course, maintaining a list of several hundred or more URLs would be unwieldy. I think for the first iteration of the solution, 100% accuracy is not critical, and 95% would be a good start. Would it be possible to easily do a minimal solution that would cover the majority of links, and then we can gradually make it more accurate my maintaining a (short) list of exceptions, or something like that? Thanks.

(In reply to comment #3)
> Thanks for the evaluation Martin. Of course, maintaining a list of several
> hundred or more URLs would be unwieldy. I think for the first iteration of the
> solution, 100% accuracy is not critical, and 95% would be a good start. Would
> it be possible to easily do a minimal solution that would cover the majority of
> links, and then we can gradually make it more accurate my maintaining a (short)
> list of exceptions, or something like that? Thanks.

Of course that it's doable, not sure which percentage what can be done easily. I wanted only mention that it's possible that the object mapping needn't be sufficient and mapping of every method/property will be necessary.