Re: [whatwg/url] Consider percent-encoding more characters in "fragment state" (#344) from Boris Zbarsky on 2017-10-03 (public-webapps-github@w3.org from October 2017)

From: Boris Zbarsky <notifications@github.com>
Date: Tue, 03 Oct 2017 14:10:04 +0000 (UTC)
To: whatwg/url <url@noreply.github.com>
Cc: Subscribed <subscribed@noreply.github.com>
Message-ID: <whatwg/url/issues/344/333853801@github.com>

> Ok. Would coalescing around Firefox's current encoding make sense?

I don't know, honestly.  At some point it would be good to involve someone from Firefox's networking team here.  Maybe @mcmanus knows who would be a good contact for that.

> maybe you could point me to the right spot?

I believe the right spot is as of today http://searchfox.org/mozilla-central/rev/a4702203522745baff21e519035b6c946b7d710d/xpcom/io/nsEscape.cpp#259-271

That table has, for each ASCII char, a bitmask indicating where it does NOT need to be escaped.  The bit for fragments is `1<<9`, so they get escaped for all the entries in the table that are < 512 (hey at least this one is easy to skim for...).

That looks like 0x00-0x20 (control chars and space), 0x22 (`"`), 0x25 (`%`), 0x3C (`<`), 0x3E (`>`), 0x60 (<code>`</code>) and anything above 0x7E.  Which I guess is what @rmisev got via his testing, good.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/whatwg/url/issues/344#issuecomment-333853801

Received on Tuesday, 3 October 2017 14:10:33 UTC