301721 - attribute normalization should not happen

Status: RESOLVED FIXED

(Core :: DOM: HTML Parser, defect)

Description

[Anne (:annevk)]

This will make us consistent with Opera and IE and will make Mozilla more
conformant to the Web Forms 2 specification:
 <http://whatwg.org/specs/web-forms/current-work/#changes>

# The attributes defined in this specification as accepting a fixed set
# of values (e.g. type) must be compared to those values using a
# case-insensitive literal comparison. Whitespace must not be trimmed
# from attribute values to make that comparison.

# Whitespace must also not be trimmed from any other attributes (e.g.
# the value attribute).

Comment 1

[Anne (:annevk)]

Attachment: testcase

Comment 2

[Jonas Sicking (:sicking) No longer reading bugmail consistently]

This sounds like a good idea as long as we check that this is something that at least IE does. Otherwise i'm sure there are plenty of people out there that depend on current behaviour.

Comment 3

[Anne (:annevk)]

The reason I filed this bug (and got Web Forms 2 changed) is because Internet Explorer does this (and Opera too). So yeah, it would probably reduce some complexity in parsing and should not do any harm.

Comment 4

[Blake Kaplan (:mrbkap) (inactive)]

What about the cases covered by bug 87894 (and the bugs linked to in that bug's first comment)? It seems that web pages do rely on our stripping leading and trailing whitespace off of at least img tags.

Comment 5

[Anne (:annevk)]

So it probably differs per attribute per element. Bah.

Comment 6

[Henri Sivonen (:hsivonen) (not reading bugmail until 2024-09-04)]

(In reply to comment #4)
> What about the cases covered by bug 87894 (and the bugs linked to in that bug's
> first comment)? It seems that web pages do rely on our stripping leading and
> trailing whitespace off of at least img tags.

URL trimming doesn't belong in the HTML parser. Complying with HTML5 now.