W3C home > Mailing lists > Public > www-dom-ts@w3.org > March 2002

DOM L1 HTML changes

From: Curt Arnold <carnold@houston.rr.com>
Date: Sat, 2 Mar 2002 17:10:17 -0600
Message-ID: <002101c1c23f$6aeb7ab0$a800a8c0@CurtMicron>
To: <www-dom-ts@w3.org>
I've added a dom1-html-matrix target.  I believe matrix production was locking up Xalan-J, at this point, I'd recommend using SAXON 6.5.1 to build the suites.

I added metadata to HTMLAnchorElement*.xml.  None of the existing tests had any <subject> elements which are used to build the test matrix.  Adding the subjects is a clerical task, mostly cutting and pasting the appropriate URL from dom1-subjects.xml to the test.

I made some global replacements in the HTML L1 tests.  

Every test defined the variable "doc" as a "Node".  While this is legal and the code does run, we know that "doc" really is a "Document" and declaring it as such removed a whole lot of unnecessary cast operations.

Changed "it's" to "its".

Changed ignoreCase="auto" to ignoreCase="false".  There may be a few cases where ignoreCase="auto" is appropriate, however it is only right in a small minority of cases and not for all but three comparisons in all those tests.

ignoreCase="auto" should only be used in those comparisons when the case of the returned value is dependent on the media type.  For example, using ignoreCase="auto" would be approriate for an assertion on an element's or attribute's name since an HTML processor should have uppercased the name and an XML processor should have preserved the existing case.
Received on Saturday, 2 March 2002 18:10:24 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 6 April 2009 12:58:46 GMT