"David Glasser" <glasser@davidglasser.net> writes:
>> --- branches/reintegrate/subversion/libsvn_ra_neon/mergeinfo.c (original)
>> +++ branches/reintegrate/subversion/libsvn_ra_neon/mergeinfo.c Thu Dec 13 20:36:27 2007
>> @@ -157,7 +157,7 @@
>> const apr_array_header_t *paths,
>> svn_revnum_t revision,
>> svn_mergeinfo_inheritance_t inherit,
>> - svn_boolean_t include_descendents, /*### TODO(reint): implement*/
>> + svn_boolean_t include_descendants,
>> apr_pool_t *pool)
>> {
>> int i, status_code;
>> @@ -185,6 +185,16 @@
>> "<S:inherit>%s"
>> "</S:inherit>",
>> svn_inheritance_to_word(inherit)));
>> +
>> + if (include_descendants)
>> + {
>> + /* Send it only if true; server will default to "no". */
>> + svn_stringbuf_appendcstr(request_body,
>> + apr_pstrdup(pool,
>> + "<S:include-descendants>yes"
>
> Why not use the #define here? (Also, no need to pstrdup a static string.)
We're not using the #defines for any of the protocol elements in
ra_neon, I think because the elements are put together in a slightly
different way there than in ra_serf. Though deploring the inconsistency
between ra_neon and ra_serf, I didn't want to introduce an even finer-
grained inconsistency within ra_neon as part of this change. (I'd be
all for cleaning this up as a separate change, though!)
Dan also brought up the pstrdup'ing. I guess you're right. I didn't
like the lifetime sloppiness of having it last longer than the other
strings in the local pool, but that seems kind of silly in retrospect.
Fixed in r28494.
> Also, does notes/webdav-protocol need to be updated?
I forgot about that file, whups. Done in r28492.
Thanks for the review, as always,
-Karl
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Sat Dec 15 01:54:23 2007