OxyGen publishes a schema for XSLT documentation which can be used to generate documentation for XSLT using OxyGen (or other tools - it's a schema after all). The XSLT documentation is useful but adds considerable bulk to any XSLT which uses it (sub-optimal for browser-based transform scenarios for example) so could be a candidate for being 'compressed' into oblivion.
Suggestion:
If the XSLT file being compressed includes the namespace xmlns:doc="http://www.oxygenxml.com/ns/doc/xsl" (or other namespace with the same URL value as the ns can be customised); and
this is not equal to the document's default namespace; and
the document declares at least one other namespace other than xsl/xslt (value http://www.w3.org/1999/XSL/Transform); then
delete elements defined in that namespace (thereby removing them and their content).
Option:
As an afterthought perhaps it would be better to accept an arg which defines a namespace (or list thereof) which could be 'safely' removed thereby putting a bit more control into the hands of the human running the compressor? This would not tie your fine and handy compressor to a particular documentation tool.
OxyGen publishes a schema for XSLT documentation which can be used to generate documentation for XSLT using OxyGen (or other tools - it's a schema after all). The XSLT documentation is useful but adds considerable bulk to any XSLT which uses it (sub-optimal for browser-based transform scenarios for example) so could be a candidate for being 'compressed' into oblivion.
Suggestion:
xmlns:doc="http://www.oxygenxml.com/ns/doc/xsl"(or other namespace with the same URL value as the ns can be customised); andhttp://www.w3.org/1999/XSL/Transform); thenOption:
As an afterthought perhaps it would be better to accept an arg which defines a namespace (or list thereof) which could be 'safely' removed thereby putting a bit more control into the hands of the human running the compressor? This would not tie your fine and handy compressor to a particular documentation tool.