Paul, have you read http://lists.xwiki.org/pipermail/users/2010-November/022514.html . The problem is not that the XHTML provided by XWord is invalid. The problem is that the syntax of the rendering markers (XML comments) has changed. Hope this helps, Marius On 12/09/2010 06:13 AM, Paul Harris wrote:
Hi again,
I'm investigating the problem with Publishing via XWord (with images), and it appears to be a problem with XWiki rather than XWord
* Edit Sandbox Test Page1 in XWord
* Clear entire document with ctrl-a, delete
* Insert an image
* Click Publish
I hacked XWord a bit so I could see what it was doing. in XWikiXMLRPCClient.cs around line 220 in method SavePageHTML(), it does this:
page.content = proxy.Convert(token, page.content, "xhtml/1.0", syntax);
lets break it down...
proxy.Convert calls: ///<summary> /// Converts a wiki source from a syntax to another syntax. ///</summary> ///<param name="token">The authentication token.</param> ///<param name="source">The content to be converted.</param> ///<param name="initialSyntaxId">The initial syntax of the source.</param> ///<param name="targetSyntaxId">The final syntax of the returned content.</param> ///<returns>The converted source.</returns> [XmlRpcMethod("confluence1.convert")] String Convert(String token, String source, String initialSyntaxId, String targetSyntaxId);
Looking on the webserver logs, I see a call to xmlrpc
I hacked XWord a little so it told me what page.content originally was... this is what was sent to xmlrpc <div class="WordSection1"> <p> <span> <!--startimage:image001.png--> <img width="236" height="158" src="image001.png" alt="demo_image.png" /> <!--stopimage--> </span> </p> </div>
so thats very simple. I've checked with another document that just contains text, and what proxy.Convert() will return is the wiki-code equivalent of the HTML that it was given.
HOWEVER, with this as the input, the xmlrpc call NEVER returns - the dialog box stays stuck, and I know it never returns as i asked a messagebox to appear after the call, which never appears.
Checking the website logs reveals a boat load of errors, I've attached them in a text file.
I can't see how XWord could do things better... its asking the xwiki for the correct syntax, and its passing valid html.
ideas?
Thanks, Paul
_______________________________________________ users mailing list [email protected] http://lists.xwiki.org/mailman/listinfo/users