Extension to Samba IDL.

[Christopher R. Hertel]

On Jun 19,  8:28pm, Elrond wrote:
:
> Hmm... You can add some documentation to the generated
> header with cpp_quote, but that doesn't realy help you...

I was looking at doing something similar (but better than) what Javadoc does.
 Javadoc, in turn, got some ideas from (how 'bout this) AmigaDOC.  The old
Amiga ROM kernel source was commented in a special format that could be parsed
out and printed as documentation.

I'm looking at yodl because I want to be able to generate docs in a variety of
formats.

> I don't know, if you've already seen "offset"-pointers.
>
> *bing* You have. You just mentioned them in your
> bit-examples. Just in DCE/RPC, they're relative to the
> start of the struct, not relative to whatever they're with
> Netbios-names.

Luke:  what's wrong with doing this?

Thing is, we'd need different syntax for relative and absolute offsets.

[reloffset] [absoffset]

??

> Well, I wanted that all to happen in the normal
> marshalling/unmarshaling-process.

Sounds good to me.

> Hmm... Yeah... Which makes up a _detail_-question: If/when
> sidlc generates the header, how should the generated header
> for bitfields look like?
> a) C-style bitfields.
> b) Just the next size of a normal storage type

Implementation of bitfields is system specific.  My memory is that some systems
will allocate a long for each bitfield.   That's okay by me for storage, but I
don't think C will do bounds checking for us so I don't see any potential gain
from option a.

> > Easier to do in the IDL code (if it's supported).
>
> You mean in the generated parser?

Right.  Which means that the IDL translater must recognize it.

> This isn't too easy at a first glance.
> Remember byte-ordering... (I think, I don't want to know
> about byte-ordering and bitfields...)

All I'm interested in is field length.  It doesn't matter how it's ordered
until you get it into a byte-oriented form.  If you had:

[y6|y7|y8|y9|yA|yB|yC|yD][x|x|y0|y1|y2|y3|y4|y5]

...then you'd be pretty certain that the data was written as a two-byte word in
the first place.  If this were the NetBIOS example, then a zero-length label
wouldn't have the second (processor order) byte anyway.

In other words, I don't think that this is a problem.

Chris -)-----

-- 
Christopher R. Hertel -)-----                   University of Minnesota
crh at nts.umn.edu              Networking and Telecommunications Services