On Sat, May 5, 2012 at 1:36 PM, Bert Huijben <bert_at_qqmail.nl> wrote:
>> -----Original Message-----
>> From: gstein_at_apache.org [mailto:gstein_at_apache.org]
>> Sent: zaterdag 5 mei 2012 18:42
>> To: commits_at_subversion.apache.org
>> Subject: svn commit: r1334446 - in /subversion/trunk/subversion/libsvn_ra_serf:
>> blame.c get_deleted_rev.c getdate.c getlocations.c getlocationsegments.c
>> getlocks.c locks.c log.c mergeinfo.c ra_serf.h replay.c update.c util.c
>>
>> Author: gstein
>> Date: Sat May 5 16:42:10 2012
>> New Revision: 1334446
>>
>> URL: http://svn.apache.org/viewvc?rev=1334446&view=rev
>> Log:
>> Switch many request handlers to use HANDLER->SLINE.CODE rather than
>> the STATUS_CODE produced by the XML parsing (which was the wrong level
>> of abstraction).
>>
>> Some handlers just don't use the status code, so it was removed. Those
>> that need it fill in a HANDLER_POOL to get the correct data from the
>> response header.
>>
>> * subversion/libsvn_ra_serf/ra_serf.h:
>> (svn_ra_serf__handler_t): clarify that HANDLER_POOL must be non-NULL
>> to get SLINE and LOCATION filled in.
>>
>> * subversion/libsvn_ra_serf/getlocationsegments.c:
>> (gls_context_t): remove STATUS_CODE
>> (svn_ra_serf__get_location_segments): switch to SLINE.CODE
>>
>> * subversion/libsvn_ra_serf/blame.c:
>> (svn_ra_serf__get_file_revs): switch to SLINE.CODE
>>
>> * subversion/libsvn_ra_serf/log.c:
>> (log_context_t): remove STATUS_CODE
>> (svn_ra_serf__get_log): switch to SLINE.CODE
>>
>> * subversion/libsvn_ra_serf/get_deleted_rev.c:
>> (svn_ra_serf__get_deleted_rev): switch to SLINE.CODE
>>
>> * subversion/libsvn_ra_serf/getlocations.c:
>> (loc_context_t): remove STATUS_CODE
>> (svn_ra_serf__get_locations): switch to SLINE.CODE
>>
>> * subversion/libsvn_ra_serf/locks.c:
>> (handle_lock): leave a todo for switch to HANDLER->SLINE
>> (svn_ra_serf__get_lock): switch to SLINE.CODE
>
> I wouldn't be surprised if you get a few XPasses when you fix the error handling here. There is a known problem here where single errors are reported multiple times.
None so far, but thanks for the heads up. I'll watch for them, as I
continue to clarify/simplify the code.
Cheers,
-g
Received on 2012-05-05 20:09:41 CEST