htmlparser-developer Mailing List for HTML Parser

It looks like headings should definitely also extend CompositeTag. I
think that's most of the essential ones now (although we still need
THEAD, TBODY and TFOOT at some point).

I've attached the file, HeadingTag.java, and also all the other tag
changes to date, to make it easier to keep track of all the changes
made so far.

PrototypicalNodeFactory has also been updated with the new tag
definitions (heading, paragraph and the definition list stuff).

I promise I'll try and get CVS upload worked out soon ;)

Best wishes

Ian Macfarlane

Here are some updates to the various table-related tags, making them
close on finding open/close TBODY, THEAD or TFOOT.

Sorry, I'm still trying to work out CVS write access.

Ian

Feature request 1291620
http://sourceforge.net/tracker/index.php?func=3Ddetail&aid=3D1291620&group_=
id=3D24399&atid=3D381402

has been fulfilled by patch 1338534:
http://sourceforge.net/tracker/index.php?func=3Ddetail&aid=3D1338534&group_=
id=3D24399&atid=3D381401

Please mark it as closed.

Ian

Forwarded again as SourceForge blocked the zip attachments on the previous =
mail.

---------- Forwarded message ----------
From: Ian Macfarlane <ian...@gm...>
Date: Oct 27, 2005 1:35 PM
Subject: More patches for HTMLParser
To: Der...@ro...
Cc: htm...@li...


Dear Derrick,

I've modified the custom definitions I had to write for p tags and
definition-list-related tags (dl,dd,dt).

Notes:

- ParagraphTag: There are a lot of tags listed as closing the P tag -
these all appear to be correct however - I have tested it using the
splitting_test.html file attached (sample testing div splitting p) and
Firefox's DOM inspector, and went through every tag in the HTML 4
specification. All the listed tag names caused the P tag to close in
Firefox.

- Definition list stuff: I did not simply add these as entries to the
Bullet and BulletList classes, as from testing (a few examples shown
in the attached list_test.html) showed that they in fact behaved
differently to normal lists (for example a dt closes a dd, and vice
versa, but a li does not close either of these, nor do they close an
li.

A question about PrototypicalNodeFactory, is there a reason that the
div, span, body, head and html tag registration in method registerTags
() are at the end rather than in alphabetical order like the other
tags?

Kind regards,

Ian Macfarlane

PS: I've tried the following command to add a file to CVS:

 cvs -d :ext:ian...@cv...:/cvsroot/htmlparser
add src/org/htmlparser/tags/ParagraphTag.java

And I get this response:

cvs add: in directory .:
cvs [add aborted]: there is no version here; do 'cvs checkout' first

I'm new to using CVS, so I'm not entirely sure what I'm doing wrong
here. Do you have any suggestions?