On Fri, 8 Jun 2001, Greg Stein wrote:
> > read on the (former) PIPE/SOCKET buckets was going to fail, it would have
> > failed earlier in the call to apr_brigade_length(). Right?
>
> Good point. (and let's also clarify to ourselves that we're just talking
> about the bytrange filter's use of partition; other partition users will
> still need to be wary of error return values)
Absolutely. No doubt about it.
> However... (and you knew that was coming :-)
;-]
> Not if the PIPE has a known length, but simply hasn't read the data yet
> (dunno if PIPE can support this; if not, then just say there is another
> bucket that does).
They do not... neither to SOCKET buckets. The general rule has always
been that any bucket that is of pre-defined length will be either a
"shared" bucket or a "simple" bucket, meaning that they can be split
without being read by simply adjusting the start/length values. The only
buckets that do not support this are buckets of indeterminate length (ie,
length == -1). But I guess that could change later (don't currently see
how, but let's just say that somebody figures out a need for such a
beast). In that case, you're right, this would be necessary. So for
future-proofness, we'll just say that it could happen. Fair enough.
> [ note: apr_brigade_partition() does not properly handle *failure* in a
> split operation; it assumes anything but ENOTIMPL is success ]
The old version did, but the patched version doesn't:
- if (apr_bucket_split(e, point) != APR_ENOTIMPL)
- return APR_BUCKET_NEXT(e);
+ if ((rv = apr_bucket_split(e, point)) != APR_ENOTIMPL) {
+ *after_point = APR_BUCKET_NEXT(e);
+ return rv;
+ }
While success and non-ENOTIMPL failure are handled by the same case, the
actual rv is returned. So if that's APR_EOF or something, it will get
passed back. *after_point is set anyway as a side effect, but that
doesn't matter. Good?
> In fact, I am going to be changing mod_dav's interface to the back-end
> repository to use bucket brigades(*). Let's say that a database backend
> inserts a bucket that will read a BLOB out of a database. It knows the
> length is N bytes, but it certainly doesn't want to read all that into
> memory right away. So when it *does* read, it could fail (e.g. the database
> connection timed out).
Good example!
--Cliff
--------------------------------------------------------------
Cliff Woolley
[EMAIL PROTECTED]
Charlottesville, VA
