[ http://issues.apache.org/jira/browse/XALANJ-2032?page=comments#action_58778 ] Santiago Pericas-Geertsen commented on XALANJ-2032: ---------------------------------------------------
Yes, the report by Alfred is accurate and correct. The basic problem is qname equality. XSLTC uses a hashtable for qname that effectively results in a "value" equality that ignores the prefix. This can be problematic together with exclude-result-prefixes as shown in the example. A possible workaround is to add the following XSLT instruction, <xsl:namespace-alias stylesheet-prefix="xhtml" result-prefix="#default"/> to ensure that a prefix alias is used. Fixing this bug will require some re-engineering of the qname allocation mechanism in the Parser. > XSLTC: Multiple namespace prefixes produce unexpected prefixes in output > ------------------------------------------------------------------------ > > Key: XALANJ-2032 > URL: http://issues.apache.org/jira/browse/XALANJ-2032 > Project: XalanJ2 > Type: Bug > Components: XSLTC > Versions: 2.6 > Reporter: Alfred Nathaniel > Assignee: Santiago Pericas-Geertsen > > The bug can be demonstrated with the following stylesheet: > <?xml version="1.0"?> > <!-- bug3.xsl --> > <xsl:stylesheet > xmlns:xsl="http://www.w3.org/1999/XSL/Transform"; > xmlns:xhtml="http://www.w3.org/1999/xhtml"; > xmlns="http://www.w3.org/1999/xhtml"; > exclude-result-prefixes="xhtml" > version="1.0"> > <xsl:template match="never"> > <xhtml:br/> > </xsl:template> > <xsl:template match="/"> > <html> > <body> > Hello<br/>world > </body> > </html> > </xsl:template> > </xsl:stylesheet> > Given any input the result produced by the Xalan interpreter is: > xalan -in bug3.xsl -xsl bug3.xsl > <html xmlns="http://www.w3.org/1999/xhtml";> > <body> > Hello<br/>world > </body> > </html> > XSLTC 2.6.0 produces a different output: > xalan -in in2.xml -xsl bug2.xsl -xsltc -v > >>>>>>> Xalan Version Xalan Java 2.6.0, <<<<<<< > <html xmlns="http://www.w3.org/1999/xhtml";> > <body> > Hello<xhtml:br xmlns:xhtml="http://www.w3.org/1999/xhtml"/>world > </body> > </html> > As XML documents both versions are equivalent that one could argue it not > really being a bug. However, contemporary browsers (Explorer6.0, > Firebox0.9.1, Mozilla1.7) ignore prefixed HTML tags, that the use of XSLTC > for HTML-to-HTML transformations is handicaped. > The problem is in org.apache.xalan.xsltc.compiler.Parser. For each namespace > there is hashtable containing all tags in the stylesheet. In the example, > <xhtml:br/> and <br/> map to the same hashtable entry but the prefix form of > the first one seen by the parser is kept. > The XSLTC code generator then uses this entry everywhere irrespective of the > prefix form used at the current position in the stylesheet. In order to avoid > the pro blem, the hashtable entries should be normalized to use a single > prefix form. > WORKAROUNDS > The safest way to avoid the problem is to avoid using multiple prefixes for > the same namespace. Note that with an appropriate <xls:import> sequence also > <xsl:template match="xhtml:br"> without any literal <xhtml:br/> can lead to > the same effect. That means for XHTML-to-XHTML transformations that all > literal tags in the stylesheet must use the prefixed notation (<xhtml:html>). > If that is too tedious for stylesheets containing a lot of literal HTML tags, > you have to rearrange the imports and template definitions that the parser > sees for all tags the non-prefixed form first. Otherwise we will have to > filter the XSLTC output to do the prefix normalization in a separate step. -- This message is automatically generated by JIRA. - If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa - If you want more information on JIRA, or have a bug to report see: http://www.atlassian.com/software/jira --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
