Closed
Bug 60511
Opened 24 years ago
Closed 23 years ago
[DOCTYPE] Should activate standards layout for ISO doctypes
Categories
(Core :: DOM: HTML Parser, defect, P2)
Core
DOM: HTML Parser
Tracking
()
RESOLVED
FIXED
mozilla1.0
People
(Reporter: hsivonen, Assigned: dbaron)
References
Details
(Whiteboard: [Hixie-P5] easy fix)
Attachments
(1 file)
(This is seems to be covered in bug 44340, but that "enhancement" is futured and I think this issue deserves to be tracked separately.) Currently, Mozilla triggers the quirks layout mode when an ISO/IEC 15445:2000 doctype declaration is used. This is a Bad Thing. ISO/IEC 15445:2000 is stricter than HTML 4 Strict. The standards layout mode should be activated for these doctype declarations: <!DOCTYPE HTML PUBLIC "ISO/IEC 15445:2000//DTD HyperText Markup Language//EN"> <!DOCTYPE HTML PUBLIC "ISO/IEC 15445:2000//DTD HTML//EN"> (regardless of a possible system identifier such as "ftp://ftp.cs.tcd.ie/isohtml/15445.dtd")
|
Reporter | |
Updated•24 years ago
|
Blocks: 34662
Keywords: mozilla0.9
ISO/IEC 15445:2000 is almost the same as ISO/IEC 15445:1999, and at present
Mozilla treats 15445:1999 as a strict mode. I think that we should not
distinguish these two DTDs. I've made a patch to fix this bug.
Index: nsParser.cpp
===================================================================
RCS file: /cvsroot/mozilla/htmlparser/src/nsParser.cpp,v
retrieving revision 3.239
diff -u -r3.239 nsParser.cpp
--- nsParser.cpp 2001/03/10 21:01:03 3.239
+++ nsParser.cpp 2001/03/15 06:37:16
@@ -752,7 +752,7 @@
else {
//now check for strict ISO/IEC OWNER...
-
if(kNotFound!=aBuffer.Find("15445:1999",PR_FALSE,theOffset,theDTDPos-theTokenizer.mOffset))
{
+
if(kNotFound!=aBuffer.Find("15445:",PR_FALSE,theOffset,theDTDPos-theTokenizer.mOffset))
{
thePublicID=eDTDMode_strict; //this ISO/IEC DTD is always
strict.
aDocType=eHTML4Text;
}
@@ -911,7 +911,7 @@
PRInt32 theStartPos=theSubIndex+5;
PRInt32 theCount=theEnd-theStartPos;
- if(kNotFound!=aBuffer.Find("ISO/IEC
15445:1999",PR_TRUE,theIndex,theEnd-theIndex)) {
+ if(kNotFound!=aBuffer.Find("ISO/IEC
15445:",PR_TRUE,theIndex,theEnd-theIndex)) {
//per spec, this DTD is always strict...
aParseMode=eDTDMode_strict;
aDocType=eHTML4Text;
This bug doesn't seem to be a higher priority compared to the rest of my m0.9.1 bugs. Moving to m0.9.2.
Target Milestone: mozilla0.9.1 → mozilla0.9.2
|
||
Updated•23 years ago
|
Whiteboard: [Hixie-P5] easy fix
|
||
Updated•23 years ago
|
Priority: P3 → P2
Moving to more realistic target - m0.9.4!
Target Milestone: mozilla0.9.3 → mozilla0.9.4
[timeless] any reason not to treat _all_ iso doctypes as strict?
<choess> timeless: I doubt it. anyone writing even HTML 2.0-equivalent with an
ISO doctype probably knows how to DTRT.
anyways, there's a patch, there's no reason for this bug to be continuously
pushed out w/o consideration. would someone please review it and address my
question [which would codewise be equivalent to suggesting:
|+ if(kNotFound!=aBuffer.Find("ISO/IEC ",PR_TRUE,theIndex,theEnd-theIndex)) {|]
|
||
Updated•23 years ago
|
|
|
||
Comment 10•23 years ago
|
||
Notes on bug: we should probably recognize anything that says "ISO/IEC 15445" for the purposes of doctype sniffing. 15445 is the ISO standard for HTML, and it gets updated periodically as HTML is updated, so a different number won't suddenly appear (unlike RFCs). While "number:year" is the currently canonical way to date a specification, "number-year" was an old syntax; I think this was changed before SGML was invented, but I don't see any harm in not bothering to check the character after "15445".
|
|
||
Comment 11•23 years ago
|
||
To clarify slightly my previous comment, I don't think we should recognize all ISO doctypes; I doubt anything with a different ISO number will come over the wire as text/html, but checking number just seems more like the Right Thing to do.
|
|
||
Comment 12•23 years ago
|
||
r=hixie on the attached patch
|
Assignee | |
Comment 13•23 years ago
|
||
I'm very concerned that the patch on this bug has been rotting here. Considering this, I think even more strongly than I did before that we need to do our DOCTYPE detection the other way around -- have a list of doctypes that trigger quirks mode, and let all unknown doctypes trigger strict mode. So I'm going to take bug 55264, and I intend to fix it for 0.9.5.
|
|
Assignee | |
Comment 15•23 years ago
|
||
Fixed with the checkin for bug 55264, 2001-09-08 11:37 PDT.
Status: NEW → RESOLVED
Closed: 23 years ago
Resolution: --- → FIXED
|
|
Reporter | |
Comment 16•22 years ago
|
||
Verified on OS X (2002040809) and Solaris (2002031500) with ISO HTML doctypes, 1999 and 2000 versions, long and short public ids, with and without a system identifier. Also verified using a fictitious public id with a future year.
You need to log in
before you can comment on or make changes to this bug.
Description
•