As a co-newbie user of Xalan, I somewhat recently tried a similar experiment. To jump to the end of the story, I'm still using Xalan, but not the xhtml2fop.xslt sheet. I managed to get the translation and PDF rendering to work, but it was not the sort of slam-dunk success I had hoped for. In the FOPS processing, there were seemingly minor details of perfectly valid XHTML that were not completely implemented, and certain major details that were implemented in a rigid manner that I couldn't live with. I can't really blame FOPS or the XSL translation. When going from a sparse representation to an information rich representation, all of the extra information has to come from somewhere, and xhtml2fop incorporates a lot of assumptions by necessity. But taking control of the fops translation was too complicated, and I gave it up in favor of the commercial Prince package, which is CSS based -- but no less fussy about valid XHTML on its input side.
Before jumping to the conclusion that Xalan is not working, I would suggest that you split your application sequence into parts. Hand compose an XHTML file and validate it with command-line TIDY. Document identification is critical (I know I'm not using the technically correct terminology here, but I refuse to memorize the XML, XSL, and XPATH specifications as I should do to make the XSL transformations in XALAN completely comprehensible). The ?xml and !DOCTYPE and namespace declarations are all critical, and if they are not exactly right XALAN will tend to obey the rules exactly and take anything it doesn't recognize (e.g. all content within tags) and spew it to the output stream. As a first step, try something that is likely to work given a valid xhtml input, such as the "identity translation" <!-- The Identity Transformation (from http://www.dpawson.co.uk/xsl/sect2/identity.html --> <xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform";> <!-- Whenever you match any node or any attribute --> <xsl:template match="node()|@*"> <!-- Copy the current node --> <xsl:copy> <!-- Including any attributes it has and any child nodes --> <xsl:apply-templates select="@*|node()"/> </xsl:copy> </xsl:template> </xsl:stylesheet> If your hand-tested xhtml content doesn't survive this parsing process, it has no chance with xhtml2fop either. Once you are over this hurdle, take on the JTIDY processing. Inspect the raw HTML in and the generated XHTML out. Does the JTIDY-generated XHTML file have all of the necessary identification and namespace tags that you learned about from your manually-composed XHTML file? Having passed two hurdles, the problem of getting the xhtml2fops translation to work should be a little easier. Before running the FOPS processing, take a close look at the XML result XALAN produces when it applies the stylesheet. It should look like a valid XML file with lots of additional "fops" tags inserted to wrap your original content. Once that is credible, give the FOPS processing and PDF output generator another try. Good luck, and I hope this message from bitter experience helps. -----Original Message----- From: John Brown [mailto:[EMAIL PROTECTED] Sent: Thursday, November 01, 2007 9:30 PM To: xalan-j-users@xml.apache.org Subject: Cannot convert XHTML to XML-FO using Xalan-J and xhtml2fso.xslstylesheet First of all, I know nothing about Xalan. The following article in the JavaWorld Forums: http://www.javaworld.com/javaworld/jw-04-2006/jw-0410-html.html explains how to convert HTML to PDF by using JTidy, Xalan and Apache FOP. The steps are as follows: 1) HTML -> XHTML uing jtidy-04aug2000r7-dev 2) XHTML -> XML-FO using xalan-j 2.7.0 3) XML-FO -> PDF using Apache fop-0.20.5 My problem is that Xalan does not seem to be working. I am using Windows XP, and I have these packages installed in c:\downloads\utils\jtidy-04aug2000r7-dev c:\downloads\utils\xalan-j_2_7_0 c:\downloads\utils\fop-0.20.5 (All binary distributions; I did not compile them myself) Java 1.6 JRE (1.6.0_03) in C:\program files\java\jre1.6.0_03 Java 1.6 JDK (1.6.0_03) in C:\program files\java\jdk1.6.0_03 xhtml2fo.xsl is an XML stylesheet for converting the output of JTidy to XML-FO. This stylesheet is supplied with the article. Following the instructions in the article, I ran the following commands: java -cp c:\downloads\utils\jtidy-04aug2000r7-dev\Tidy.jar -asxml \ hello.html> temp.xml java -cp \ c:\downloads\utils\xalan-j_2_7_0\xalan.jar\ ;c:\downloads\utils\xalan-j_2_7_0\xercesImpl.jar\ ;c:\downloads\utils\xalan-j_2_7_0\xml-apis.jar\ ;c:\downloads\utils\xalan-j_2_7_0\serializer.jar\ org.apache.xalan.xslt.Process -TT -TG -TS -TTC -IN temp.xml\ -XSL xhtml2fo.xsl -OUT temp.fo c:\downloads\utils\fop-0.20.5\fop.bat temp.fo hello.pdf Of course, I did not really use the "\" character. Each command is on one line, with no unnecessary white space. JTidy produces the following XML file: Hello World Hello World! This seems to be OK. The console output of xalan is: null Line #0, Column #0: template match='/' file:///C:/Downloads/Utils/html2pdf/xhtml2fo.xsl Line #0, Column #0: apply-templates Selected source node '#document', at file 'file:///C:/Downloads/Utils/html2pdf/temp.xml', line #-1, column #-1 (default root rule) apply-templates, select='null': 10001: html null Line #0, Column #0: template match='*' file:///C:/Downloads/Utils/html2pdf/xhtml2fo.xsl Line #0, Column #0: apply-templates Selected source node 'html', at file 'file:///C:/Downloads/Utils/html2pdf/temp.xml', line #-1, column #-1 (default rule) apply-templates, select='null': 10004: #text 10005: head 1000e: #text 1000f: body 10014: #text STARTDOCUMENT CHARACTERS: null Line #0, Column #0: template match='*' file:///C:/Downloads/Utils/html2pdf/xhtml2fo.xsl Line #0, Column #0: apply-templates Selected source node 'head', at file 'file:///C:/Downloads/Utils/html2pdf/temp.xml', line #-1, column #-1 (default rule) apply-templates, select='null': 10006: #text 10007: meta 1000a: #text 1000b: title 1000d: #text CHARACTERS: null Line #0, Column #0: template match='*' file:///C:/Downloads/Utils/html2pdf/xhtml2fo.xsl Line #0, Column #0: apply-templates Selected source node 'meta', at file 'file:///C:/Downloads/Utils/html2pdf/temp.xml', line #-1, column #-1 (default rule) apply-templates, select='null': [empty node list] CHARACTERS: null Line #0, Column #0: template match='*' file:///C:/Downloads/Utils/html2pdf/xhtml2fo.xsl Line #0, Column #0: apply-templates Selected source node 'title', at file 'file:///C:/Downloads/Utils/html2pdf/temp.xml', line #-1, column #-1 (default rule) apply-templates, select='null': 1000c: #text CHARACTERS: Hello World CHARACTERS: CHARACTERS: null Line #0, Column #0: template match='*' file:///C:/Downloads/Utils/html2pdf/xhtml2fo.xsl Line #0, Column #0: apply-templates Selected source node 'body', at file 'file:///C:/Downloads/Utils/html2pdf/temp.xml', line #-1, column #-1 (default rule) apply-templates, select='null': 10010: #text 10011: p 10013: #text CHARACTERS: null Line #0, Column #0: template match='*' file:///C:/Downloads/Utils/html2pdf/xhtml2fo.xsl Line #0, Column #0: apply-templates Selected source node 'p', at file 'file:///C:/Downloads/Utils/html2pdf/temp.xml', line #-1, column #-1 (default rule) apply-templates, select='null': 10012: #text CHARACTERS: Hello World! CHARACTERS: CHARACTERS: ENDDOCUMENT I do not understand any of this, but all those lines with line #-1, column #-1 cannot be good. The resulting file is: Hello World Hello World! This is obviously wrong. The article supplied a file Html2Pdf.java, which is supposed to be the equivalent of the sequence of commands. I compiled it, and it works. What am I doing wrong? _________________________________________________________________ Help yourself to FREE treats served up daily at the Messenger Café. Stop by today. http://www.cafemessenger.com/info/info_sweetstuff2.html?ocid=TXT_TAGLM_OctWL tagline
