483660 - Consider not setting gEscapeUTF8 to true

Status: RESOLVED WONTFIX

(Core :: Networking, defect, P3)

Description

[Boris Zbarsky [:bzbarsky]]

Right now, nsStandardURL escapes all non-ASCII bytes in its internal representation, as far as I can tell.  It certainly does so for refs, which causes issues like bug 135309.

Is there a reason we're not storing in IRI form internally, escaping to URI as needed when someone actually needs an ASCII spec?  We have GetAsciiSpec anyway...

Comment 1

[Boris Zbarsky [:bzbarsky]]

That said, is flipping that pref basically all we need to get IRI behavior?

Comment 2

[Patrick McManus [:mcmanus]]

I think we're really not pursuing iri at this point. reopen if I've got that wrong

Comment 3

[Boris Zbarsky [:bzbarsky]]

This blocks observable bugs in our HTML support, so if we're not going to change necko here we need to figure out some other approach...  I still think fixing this in necko (via aligning with the URL spec, not IRI) is the right thing.

Comment 4

[Firefox Bug Husbandry Bot]

Bulk change to priority: https://bugzilla.mozilla.org/show_bug.cgi?id=1399258

Comment 5

[Firefox Bug Husbandry Bot]

Bulk change to priority: https://bugzilla.mozilla.org/show_bug.cgi?id=1399258

Comment 6

[Anne (:annevk)]

How exactly are we not aligned with the URL Standard here? The URL Standard requires UTF-8 percent encoding of non-ASCII input (except for the query, where we use the document encoding to do so). I believe Gecko/Necko is aligned with that.

Comment 7

[Boris Zbarsky [:bzbarsky]]

> How exactly are we not aligned with the URL Standard here?

No idea, sorry.  I have been tracking neither the URL standard nor Necko on this issue.  I expect Valentin knows where things stand.

Comment 8

[Valentin Gosu [:valentin] (he/him)]

This pref was removed in Bug 1321705.
What Anne said in comment 6 is what we do.