Re: XHTML: <?xml?> shouldn't be required or a workaround needed for Microsoft stupidity
Jacob Kjome <[email protected]> Tue, 30 Sep 2003 12:47:11 -0500
| Newsgroups | gmane.comp.java.enhydra.xmlc |
|---|---|
| Message-ID | <[email protected]> |
Hi Petr, At 05:42 PM 9/30/2003 +0200, you wrote: >Hi, > >in short: we need the Xerces parser in XMLC to handle XHTML files >without the <?xml?> declaration on the first line. Reason: IE6 (and also >Opera7) does not switch to standards compliant CSS mode if the ><!DOCTYPE> is not the very first line. So lame, I know. But we need to >live with that somehow. > >Some very relevant links: >http://msdn.microsoft.com/library/default.asp?url=/library/en-us/dnie60/html/cssenhancements.asp >http://gutfeldt.ch/matthias/articles/doctypeswitch.html > >Solution to this problem has two parts: first, it should be possible to >XMLCompile a document without the <?xml?> declaration. Yeah, I guess I can agree with this for XHTML1.0 since XHTML1.0 documents don't require the XML header. Do XHTML1.1 documents require the XML header? I think they do, but I haven't verified this. >Second, the >compiled java code should offer a switch to either enable or disable >generating the <?xml?> at the serialization time (writeDOM). It's >obvious that the existing XHTMLCompatibility() function would disable >generating this <?xml?> header.. Have you tried OutputOptions.setOmitXMLHeader(true)? This is certainly a candidate for automatic setting via OutputOptions.setEnableXHTMLCompatibility(true). I'd certainly like to hear opinions from others on this, but I'm not opposed to doing it. >Please note that http://www.w3.org/TR/xhtml1/#strict says "An XML >declaration is not required in all XML documents;" so I hope XMLC can be >fixed to not reply with: > >Error: Document appears to be an HTML document; the xerces parser only >supports XML (does the document start with `<?xml ... ?>'?) Yes, I hope it won't be too difficult as well, but I guess I'm less concerned about the mockups being required to contain the XML header when it is so easily stripped upon serialization using OutputOptions. If this is a feature that you need very soon, I suggest digging into the compiler and seeing if you can come up with a patch to change this behavior for XHTML1.0 documents. I may get to this, but it won't be a priority for me since OutputOptions can already take care of this at runtime with zero changes to XMLC. Jake >Thanks > >Petr > > >_______________________________________________ >XMLC mailing list >[email protected] >http://www.enhydra.org/mailman/listinfo.cgi/xmlc