36872 - Compatibility mode testing should be reviewed

Status: VERIFIED FIXED

(Core :: DOM: HTML Parser, defect, P3)

Description

[Masatoshi Kimura [:emk]]

The mode testing code such as "mode != eParseMode_noquirks" is broken now since 
new eParseMode_strict is erroneously treated as quirks.
They should be changed to "mode == eParseMode_quirks" or "mode != 
eParseMode_strict".
It is case-by-case which change is better.
Similarly, "mode == eParseMode_noquirks" should be changed to "mode != 
eParseMode_quirks" or "mode == eParseMode_strict".

Comment 1

[rickg]

Do you know of a specific problem, or this is just "being careful"?

Comment 2

[Masatoshi Kimura [:emk]]

Attachment: quirks mode testcase

Comment 3

[Masatoshi Kimura [:emk]]

Attachment: another quirks mode testcase

Comment 4

[Masatoshi Kimura [:emk]]

Attachment: strict mode testcase

Comment 5

[Masatoshi Kimura [:emk]]

Currently, only (nonexistent) "HTML 6.0" (or later) DOCTYPES are handled in 
noquirks, so I couldn't attach a noquirks testcase.
HTML 4.0 (Strict) is handled in strict, but comments are parsed as if it is 
handled in quirks.

Comment 6

[Masatoshi Kimura [:emk]]

Here is all noquirks/strict problems I found:
http://www3.airnet.ne.jp/emk/moz/doctypes.cgi

Comment 7

[rickg]

*** Bug 36995 has been marked as a duplicate of this bug. ***

Comment 8

[rickg]

Fixed in my tree, awaiting chance to land.

Comment 9

[rickg]

Fixed in my tree.

Comment 10

[rickg]

Fixed by changing _noquirks to _strict.

Comment 11

[rickg]

Fixed by changing _noquirks to _strict.

Comment 12

[Masatoshi Kimura [:emk]]

Verified using 2000050408 nightly build on Windows 2000.

Comment 13

[David Baron :dbaron: (⌚️UTC-5, no longer working on Mozilla)]

Reopening.  My understanding was that the difference between noquirks and strict
was only in the parser.  Therefore the tests should have been changed from
== noquirks    to    != quirks
!= noquirks    to    == quirks

The current code looks very wrong.

Comment 14

[rickg]

Um, I don't think so. David if you have a particular concern, raise it. But the 
code was changed to reflect quirks vs. strict, and no_quirks has been removed.

Comment 15

[Pierre Saslawsky]

David: in order to fix a regression on bug 18817, I just checked in the 
following change per suggestion of [e-mail].

/layout/html/document/src/nsHTMLContentSink.cpp, line 3715:
    if (eDTDMode_noquirks== mode) {
should be:
    if (eDTDMode_strict == mode) {

In that case at least, the fix is correct. I don't think that 
changing "==noquirks" to "!=quirks" everywhere is the right thing to do.

Comment 16

[bsharma]

updated qa contact.

Comment 17

[bsharma]

Verified on 2001-07-19-branch on WinNT

The attached test cases loads fine.