[xmlroff] Nearly ready for xmlroff 0.5.1

[Emmanuel Fleury]

Tony Graham wrote:
> 
> Mostly.  Tests should be added when something new is implemented, and
> from them on they are used to test for regressions.  I've also used the
> 'testing' module to compare xmlroff output against output from other
> processors.
> 
> I actually had three changes in my test results, including two segfaults
> that were cleared up thanks to [1].

Ok.

> I added ticket #39 about the PDF links that won't resolve.  It should be
> a simple XSLT fix since it would be a variation on the existing
> stylesheet logic that emits "No output" text when there is no PDF.
> 
> I'm afraid that the README is the best info available.  There's also
> some text about it in my XTech 2007 paper [3].
> 
> Of course you can ask questions about it here.

Thanks. I'll probably do that if I don't understand something. :)

> There are some tickets about the 'testing' module [2].  Feel free to add
> tickets about issues that you find or, if you are so inclined, fix any
> of the open tickets.

I might. I'm quite interested in xmlroff (not only for the testing).

Regards
-- 
Emmanuel Fleury

Nothing I do is my fault.
  -- Calvin & Hobbes (Bill Waterson)