RE: DOCBOOK-APPS: Keeping entities in a transformation

[David Cramer]

I've always wanted to know if there's a right way to do that myself. I've resorted to using sed to obfuscate the entities before processing, then unobfuscating them afterwards. For each file,   sed "s/&\(\w*\.*\w*\);/%%\1%%/g" $file > $file.tmp1.xml   the run the xsl on the file, then   sed "s/%%\(\w*\.*\w*\)%%/\&\1;/g" $file.tmp2.xml > $file   It seems like such a hack--there must be a better way.   David -----Original Message----- From: Jeff Beal [mailto:[EMAIL PROTECTED]] Sent: Thursday, May 02, 2002 10:48 AM To: DocBook Mailing List (E-mail) Subject: DOCBOOK-APPS: Keeping entities in a transformation I’m working an an XSL script to clean up my DocBook XML.  I’m running into problems because all of the entities are declared in a top-level driver file, and the content resides in about 3,000 external entities, which do not include DOCTYPE declarations.  I need to maintain our current file structure and naming scheme (which unfortunately does not correspond directly to the XML.).   How can I get the XSL process to “ignore” undeclared entities and copy the reference through to the result tree?   Jeff Beal Ansys, Inc. (724)514-3150 [EMAIL PROTECTED]