Try <http://stx.sourceforge.net/documents/spec-stx-20040701.html> http://stx.sourceforge.net/documents/spec-stx-20070427.html It looks like the usual intimidating W3C documents, but isn't as bad as it first seems. The main difference is that you don't have 'apply-templates' which would be a recursive rather than sequential operation. Your options are typically to 'process-children', 'process-attributes', 'process-siblings' -- these are things that you can reference from a single location in the document while the rest is not in memory.
-----Original Message----- From: Mikael Jansson [mailto:[EMAIL PROTECTED] Sent: Friday, November 21, 2008 3:21 PM To: xalan-j-users@xml.apache.org Subject: Fwd: Transforming huge XML-files - 3-4GB "Have you heard of stx? Its a language similar in style to XSLT but it treats the input document as a stream of SAX events instead of loading the entire document into memory. We used this to process documents of 6Gb and more. The java implementation is called Joost..." OK, yeah so this looks nice. But I can't find a good tutorial on how to convert xsl into stx documents. Could it be so simple as to just change this: <xsl:stylesheet version="1.0"> <xsl:output method="text"/> <xsl:template match="posttyp"> to <stx:stylesheet version="1.0"> <stx:output method="text"/> <stx:template match="posttyp"> ... and so on? I only use some of the most basic commands: xsl:stylesheet xsl:output method xsl:template match xsl:if test xsl:apply-templates xsl:value-of select and thats it. Where can I find out what the corresponding stx commands (if not the same) are? Micke Jansson ---------- Forwarded message ---------- From: Andy Chambers < [EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]> > Date: Fri, Nov 21, 2008 at 23:31 Subject: Re: Transforming huge XML-files - 3-4GB To: Mikael Jansson < [EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]> > On Thu, Nov 20, 2008 at 8:31 PM, Mikael Jansson < [EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]> > wrote: > Hi! I've already posted this, but it does not show up in the history and > there was no response so I'll post it again. Sorry if it's a double post... > > I'm trying to transform a set of XML-files into SQL-code. It works just fine > with XSL no mater what components I use - Stream or SAX or what ever. Have you heard of stx? Its a language similar in style to XSLT but it treats the input document as a stream of SAX events instead of loading the entire document into memory. We used this to process documents of 6Gb and more. The java implementation is called Joost. What's nice is that you can "call out" to a real XSLT processor by passing fragments of the input document to xalan or saxon if you don't want to leave the comfort of programming in XSLT. Our project sounds like it was maybe pretty similar to yours. Loading huge wads of XML into a database. Word of warning though. You're probably going to need every ounce of performance you can get so make sure you're using prepared statements and you probably want to check out other XSLT processors against your intended input. Also consider whether its possible to avoid XML entirely. If the database has a bulk-loader its almost certainly going to be quicker than parsing the XML and using Java to execute the SQL statements (even if you're doing prepared statements) -- Andy
