On Fri, Aug 08, 2003 at 12:23:08AM -0400, Sergey A. Lipnevich wrote:
> properties support, I'd greatly appreciate an intermediate review so
> that I know I'm moving forward. Could somebody (MBK?) please review the
> neon-0.24 branch changes, at least briefly?
Some quick things (in addition to kfogel's review):
I cleaned up the long lines before my review; you might want to update
now so as not to generate conflicts later.
davautocheck.sh belongs in /tools/dev, not /. I'm not convinced that
it should even go there; I suspect that most people who do a lot of
make davcheck have this problem solved already.
In the shim handlers, you use a single accumulator in the baton.
This is safe for XML documents iff they only contain cdata in "leaf"
elements. I think that this is the case for ra_dav, but I could easily
be wrong. Anyway, it might make sense to add some error detection to
the shims so that if non-leaf, non-whitespace cdata is encountered, we throw
an error, instead of silently discarding it.
Also, it might be nicer to ignore the whitespace on the way into the cdata
handler, rather than stripping it later in shim_endelem. Either way, you
need to pay attention to the elem flags (XML_CDATA and XML_COLLECT)
somewhere. This will probably involve introducing some state to the
shim_baton, and some state logic to the start/cdata/end handlers.
It is likely that this is why the tests are failing. I don't know,
and haven't looked any further. You might try fixing this first, and
then debugging more.
--ben
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Sun Aug 10 22:22:53 2003