18 Aug
2008
18 Aug
'08
4:37 a.m.
Hi Vincent,
On Mon, Aug 18, 2008 at 2:59 AM, Vincent Massol <vincent(a)massol.net> wrote:
Hi Wang,
I have some questions/remarks on the commit below:
1) It's hard to see what you're modifying from the default tag infos
(in default.xml). I think a better solution might be to extend
DefaultTagProvider and use the constructor accepting an info provider,
see http://htmlcleaner.sourceforge.net/doc/org/htmlcleaner
HtmlCleaner.html#HtmlCleaner(org.htmlcleaner.ITagInfoProvider)
. In this manner we just extend the default and not duplicate it.
WDYT? Is this possible? Are you adding tags or redefining default one?
I have seen this DefaultTagProvider, but I don't think it can provide
what I want.
I want to replace some tags with others. Such like replace <p> in <li>
to <span>, replace deprecated tags with others. Like replace <s> to
<del>. But the default.xml can define this.
2) The following code is not required:
I have tried
props.setOmitHtmlEnvelope(true), but there are some
errors. If input is
<html>
<head/>
<body>
<p>p1</p>
<p>p2</p>
</body>
</html>
the output TagNode is not <p>p1</p><p>p2</p> but only
<p>p2</p>. So I
have to use removeHead() to remove the head tag.
+ private void removeHead(TagNode node)
+ {
+ logger.debug("remove the head tag of the html");
+ TagNode head = node.findElementByName("head", true);
+ head.removeFromTree();
+ }
instead you can just use props.setOmitHtmlEnvelope(true);
3) Could you explain why we need convertAllAttribute2Lowercase?
In xhtml, all the attributes value should be lowercase. And some
attributes value in the conversion result of openoffice is uppercase,
like algin="CENTER". I think lowercase will be better. But if
xhtmlparser could handler uppercase attributes value, I will remove
this method.
4) Could you refactor your code to use the new
HTMLCleaner code I have
put in xwiki-xml module?
I will do it today.
Feel free to provide patches for CleaningFilters.
[snip]
--
Thanks
Wang Ning