Re: [avr-libc-dev] New device support

[Joerg Wunsch]

As Eric Weddington wrote:

> > I did want to get a feeling for XSLT, but I'm still not really
> > convinced about it.  Btw., I did make an initial proof-of-concept
> > XSLT "style sheet" to convert the XML to header files, mainly in
> > order to demonstrate to Reinhard it's doable, but this
> > implementation rather let me realize the Python version is really
> > the better way to go. ;-)

> Can you describe your reasoning?

The bit names being far away from the register names in XML is one
thing.  Another thing is to summarize multiple registers into 16-bit
pseudo-registers, probably based on some kind of outside information
because there are no hints about that in the XML.  For example, you
can summarize UBRR1L and UBRR1H into a UBRR1 pseudo-register on the
ATmega128, but you cannot summarize UBRR0L and UBRR0H because they are
not adjacent.  The closer you look, the more gotchas like these you
can find which are fairly easy to solve in a programming language
(just check for common register names ending up in `L' and `H', and
being adjacent in address space) but close to impossible in XSLT.

You cannot base an XSLT transformation on `outside information', it
can only read a single XML file at a time.  That's part of its
concept.

-- 
cheers, J"org               .-.-.   --... ...--   -.. .  DL8DTL

http://www.sax.de/~joerg/                        NIC: JW11-RIPE
Never trust an operating system you don't have sources for. ;-)