I really think most of those warnings aren't a result of your code,
but a result of our CFLAGS, which turns on every single warning
GCC can throw ("-Wall -W"); plus, the makefiles might be invoking
the C++ compiler which would be even more strict if you've got EGCS
installed.
In other words, your code works well where it knows it will be--in
C land.
> p.s. i put up a temporary development version of mswordview at
> http://www.csn.ul.ie/~caolan/publink/mswordview/development/
> which should be a better base to put an importer on top of. comes with a
> libwv.a, and has a sample usage in wvText which should make importing word
> files easier. Im putting some work into it at the moment, so it only is
> guaranteed to work with fullsave files, and then only outputs the main body
> text with <p></p> at each para boundary, though the correct paragraph
> properties for each paragraph are correctly assembled at each para start
> boundary.
Great, I heard from Justin about this new library-like interface
and I'm planning on starting some work with it today. It looks
like it's just about exactly the right level to use as a low-level
importer tool.
> I'm going to try to have wv support abiword natively, with the advantage
> that fixes to the main tree of wv can automatically benefit abiword, i'll
> let you know if it works out.
That's great. I'd like there to be a library interface (which looks
like
it will be libwv) which all sorts of programs can use to finally get
this document format usable across many systems. With a nice neutral
parser library doing all the dirty work inside the streams, everyone
else can benefit from everyone else's contributions.
-- Shaw Terwilliger