You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by Ivan Zhakov <iv...@visualsvn.com> on 2015/08/28 19:06:47 UTC
Re: svn commit: r1555133 - in /subversion/trunk: ./
subversion/include/ subversion/include/private/ subversion/libsvn_client/
subversion/libsvn_delta/ subversion/libsvn_fs_fs/ subversion/libsvn_fs_x/
subversion/libsvn_ra/ subversion/libsvn_ra_serf/ subvers...
On 3 January 2014 at 19:11, <rh...@apache.org> wrote:
> Author: rhuijben
> Date: Fri Jan 3 16:10:59 2014
> New Revision: 1555133
>
> URL: http://svn.apache.org/r1555133
> Log:
> Get rid of the ever increasing number of private apis in svn_sorts.h by adding
> a new private/svn_sorts_private.h file with the private parts and updating the
> needed #include-s.
>
[..]
> Modified: subversion/trunk/subversion/libsvn_ra_serf/util.c
> URL: http://svn.apache.org/viewvc/subversion/trunk/subversion/libsvn_ra_serf/util.c?rev=1555133&r1=1555132&r2=1555133&view=diff
> ==============================================================================
> --- subversion/trunk/subversion/libsvn_ra_serf/util.c (original)
> +++ subversion/trunk/subversion/libsvn_ra_serf/util.c Fri Jan 3 16:10:59 2014
> @@ -2354,7 +2354,7 @@ expat_response_handler(serf_request_t *r
> apr_status_t status;
> const char *data;
> apr_size_t len;
> - int expat_status;
> + svn_error_t *err;
> svn_boolean_t at_eof = FALSE;
>
> status = serf_bucket_read(response, PARSE_CHUNK_SIZE, &data, &len);
> @@ -2373,12 +2373,11 @@ expat_response_handler(serf_request_t *r
>
> /* ### should we have an IGNORE_ERRORS flag like the v1 parser? */
>
> - expat_status = XML_Parse(ectx->parser, data, (int)len,
> - at_eof /* isFinal */);
> + err = parse_xml(ectx->parser, data, len, at_eof /* isFinal */);
>
> - if (at_eof
> - || ectx->inner_error
> - || expat_status != XML_STATUS_OK)
> + err = svn_error_compose_create(ectx->inner_error, err);
> +
> + if (at_eof || err)
> {
> /* Release xml parser state/tables. */
> apr_pool_cleanup_run(ectx->cleanup_pool, &ectx->parser,
> @@ -2392,14 +2391,7 @@ expat_response_handler(serf_request_t *r
>
> If an error is not present, THEN we go ahead and look for parsing
> errors. */
> - SVN_ERR(ectx->inner_error);
> - if (expat_status != XML_STATUS_OK)
> - return svn_error_createf(SVN_ERR_XML_MALFORMED, NULL,
> - _("The %s response contains invalid XML"
> - " (%d %s)"),
> - ectx->handler->method,
> - ectx->handler->sline.code,
> - ectx->handler->sline.reason);
> + SVN_ERR(err);
>
> /* The parsing went fine. What has the bucket told us? */
>
These changes in subversion/libsvn_ra_serf/util.c looks like unrelated
and undocumented. Is not it?
--
Ivan Zhakov
Re: svn commit: r1555133 - in /subversion/trunk: ./
subversion/include/ subversion/include/private/ subversion/libsvn_client/
subversion/libsvn_delta/ subversion/libsvn_fs_fs/ subversion/libsvn_fs_x/
subversion/libsvn_ra/ subversion/libsvn_ra_serf/ subvers...
Posted by Julian Foad <ju...@btopenworld.com>.
Bert, ping?
- Julian
On 28 August 2015, Ivan Zhakov wrote:
> On 3 January 2014 at 19:11, <rh...@apache.org> wrote:
>> Author: rhuijben
>> Date: Fri Jan 3 16:10:59 2014
>> New Revision: 1555133
>>
>> URL: http://svn.apache.org/r1555133
>> Log:
>> Get rid of the ever increasing number of private apis in svn_sorts.h by adding
>> a new private/svn_sorts_private.h file with the private parts and updating the
>> needed #include-s.
>>
> [..]
>
>
>> Modified: subversion/trunk/subversion/libsvn_ra_serf/util.c
>> URL: http://svn.apache.org/viewvc/subversion/trunk/subversion/libsvn_ra_serf/util.c?rev=1555133&r1=1555132&r2=1555133&view=diff
>> ==============================================================================
>> --- subversion/trunk/subversion/libsvn_ra_serf/util.c (original)
>> +++ subversion/trunk/subversion/libsvn_ra_serf/util.c Fri Jan 3 16:10:59 2014
>> @@ -2354,7 +2354,7 @@ expat_response_handler(serf_request_t *r
>> apr_status_t status;
>> const char *data;
>> apr_size_t len;
>> - int expat_status;
>> + svn_error_t *err;
>> svn_boolean_t at_eof = FALSE;
>>
>> status = serf_bucket_read(response, PARSE_CHUNK_SIZE, &data, &len);
>> @@ -2373,12 +2373,11 @@ expat_response_handler(serf_request_t *r
>>
>> /* ### should we have an IGNORE_ERRORS flag like the v1 parser? */
>>
>> - expat_status = XML_Parse(ectx->parser, data, (int)len,
>> - at_eof /* isFinal */);
>> + err = parse_xml(ectx->parser, data, len, at_eof /* isFinal */);
>>
>> - if (at_eof
>> - || ectx->inner_error
>> - || expat_status != XML_STATUS_OK)
>> + err = svn_error_compose_create(ectx->inner_error, err);
>> +
>> + if (at_eof || err)
>> {
>> /* Release xml parser state/tables. */
>> apr_pool_cleanup_run(ectx->cleanup_pool, &ectx->parser,
>> @@ -2392,14 +2391,7 @@ expat_response_handler(serf_request_t *r
>>
>> If an error is not present, THEN we go ahead and look for parsing
>> errors. */
>> - SVN_ERR(ectx->inner_error);
>> - if (expat_status != XML_STATUS_OK)
>> - return svn_error_createf(SVN_ERR_XML_MALFORMED, NULL,
>> - _("The %s response contains invalid XML"
>> - " (%d %s)"),
>> - ectx->handler->method,
>> - ectx->handler->sline.code,
>> - ectx->handler->sline.reason);
>> + SVN_ERR(err);
>>
>> /* The parsing went fine. What has the bucket told us? */
>>
> These changes in subversion/libsvn_ra_serf/util.c looks like unrelated
> and undocumented. Is not it?