]>
granicus.if.org Git - php/log
Christian Stocker [Mon, 5 Sep 2011 08:27:47 +0000 (08:27 +0000)]
Revert r316043, the tests are different on purpose
Stanislav Malyshev [Mon, 5 Sep 2011 00:49:03 +0000 (00:49 +0000)]
fix tests
Gustavo André dos Santos Lopes [Mon, 5 Sep 2011 00:39:39 +0000 (00:39 +0000)]
- Fixed bad xor in signed types due to integer promotion.
- Replaced undefined signed overflow with char -> unsigned char conversion.
Gustavo André dos Santos Lopes [Sun, 4 Sep 2011 23:00:47 +0000 (23:00 +0000)]
- Fixed bug #55576: Cannot conditionally move uploaded file without race
condition.
Gustavo André dos Santos Lopes [Sun, 4 Sep 2011 22:36:33 +0000 (22:36 +0000)]
- Fixed borked refactoring in r307437 (using SUCCESS/FAILURE return instead of
out parameter).
- Fixed signature of php_stream_copy_to_stream_ex to return int in 5.4/trunk
instead of size_t, as the function only returns SUCCESS/FAILURE.
Pierre Joye [Sun, 4 Sep 2011 22:21:58 +0000 (22:21 +0000)]
- add skipif
Pierre Joye [Sun, 4 Sep 2011 22:12:42 +0000 (22:12 +0000)]
- force LF eol
Pierre Joye [Sun, 4 Sep 2011 22:04:42 +0000 (22:04 +0000)]
- skip on win, TZ cannot be set using putenv at runtime
Pierre Joye [Sun, 4 Sep 2011 21:59:27 +0000 (21:59 +0000)]
- skip on win, TZs differ with strftime
Pierre Joye [Sun, 4 Sep 2011 21:48:22 +0000 (21:48 +0000)]
- unify warning between win and unix and enable test
Pierre Joye [Sun, 4 Sep 2011 21:40:22 +0000 (21:40 +0000)]
- fix test on win, does not use putenv to set the TZ but date_default_timezone_set (should be the case on all platforms actually...)
Stanislav Malyshev [Sun, 4 Sep 2011 18:50:05 +0000 (18:50 +0000)]
fix bug 55562 - make substr always return rest of the string if length is too long
Pierre Joye [Sun, 4 Sep 2011 16:33:31 +0000 (16:33 +0000)]
- wrong variable used (ended as non initiazed usage)
Ferenc Kovacs [Sun, 4 Sep 2011 15:02:08 +0000 (15:02 +0000)]
passing an invalid session.upload_progress.freq always generates an additional startup error
Ferenc Kovacs [Sun, 4 Sep 2011 14:28:52 +0000 (14:28 +0000)]
passing an invalid session.upload_progress.freq always generates an additional startup error
Tjerk Anne Meesters [Sun, 4 Sep 2011 14:21:27 +0000 (14:21 +0000)]
Patch to run two more mysql test cases that would usually be skipped.
Replaced skipifdefaultconnectionfailure.inc by three additional ini_set() statements in connect.inc for the default connection settings
Patch also includes a typo in connect.inc concerning mysql.default_socket
Stanislav Malyshev [Sun, 4 Sep 2011 05:00:30 +0000 (05:00 +0000)]
fix test
Stanislav Malyshev [Sun, 4 Sep 2011 04:57:01 +0000 (04:57 +0000)]
test fixes
Christopher Jones [Fri, 2 Sep 2011 22:19:03 +0000 (22:19 +0000)]
gcov test fixes
Ferenc Kovacs [Fri, 2 Sep 2011 20:28:39 +0000 (20:28 +0000)]
pass an absolute path for the php binary to proc_open, without this the test would fail if ran with a relative TEST_DEV_EXECUTABLE
Shein Alexey [Fri, 2 Sep 2011 18:01:20 +0000 (18:01 +0000)]
Synchronized fixed tests from 5.3 branch.
Shein Alexey [Fri, 2 Sep 2011 17:41:39 +0000 (17:41 +0000)]
Removed old (< 5.3) version of test and cleaned up the actual version.
Andrey Hristov [Fri, 2 Sep 2011 13:53:41 +0000 (13:53 +0000)]
Fix for Bug #54158 MYSQLND + PDO MySQL requires #define MYSQL_OPT_LOCAL_INFILE
and a bunch of other small preprocessor fixes
Ulf Wendel [Fri, 2 Sep 2011 13:35:17 +0000 (13:35 +0000)]
Coverage for bug#54704, of which I think it is bogus, thus no dedicated test.
Andrey Hristov [Fri, 2 Sep 2011 12:59:32 +0000 (12:59 +0000)]
fix message
Andrey Hristov [Fri, 2 Sep 2011 11:22:41 +0000 (11:22 +0000)]
function rename to comply with the schema of other callbacks
Ulf Wendel [Fri, 2 Sep 2011 11:06:51 +0000 (11:06 +0000)]
Hopefully, this is an even better way to check for InnoDB support as of MySQL 5.6.1
Ulf Wendel [Fri, 2 Sep 2011 10:38:51 +0000 (10:38 +0000)]
As of MySQL 5.6.1 the server variable have_innodb is no more. Updating InnoDB check...
Boris Lytochkin [Fri, 2 Sep 2011 10:13:30 +0000 (10:13 +0000)]
merge from trunk
added SNMPException class, enabling ability to throw exceptions
when a known SNMP error has occured
FR #55542
Ulf Wendel [Fri, 2 Sep 2011 09:27:23 +0000 (09:27 +0000)]
MySQL 5.6 update
Boris Lytochkin [Fri, 2 Sep 2011 08:09:23 +0000 (08:09 +0000)]
merge from trunk:
remove php_snmp_get_ce()
Gustavo André dos Santos Lopes [Fri, 2 Sep 2011 05:14:32 +0000 (05:14 +0000)]
- Use our external entity loader only in the context of a PHP request.
Christopher Jones [Thu, 1 Sep 2011 19:27:14 +0000 (19:27 +0000)]
Unmatched %s was still diffing on Linux. Now it matches the whitespace at a minimum
Daniel Convissor [Thu, 1 Sep 2011 19:07:13 +0000 (19:07 +0000)]
Make test work with database other than "test".
Christopher Jones [Thu, 1 Sep 2011 19:01:37 +0000 (19:01 +0000)]
Prevent the magic quotes fatal error causing a diff
Ulf Wendel [Thu, 1 Sep 2011 16:35:13 +0000 (16:35 +0000)]
Check for LOAD DATA support during SKIPIF
Ulf Wendel [Thu, 1 Sep 2011 15:45:47 +0000 (15:45 +0000)]
Fixing error handling. Skip if load data is not allowed
Ulf Wendel [Thu, 1 Sep 2011 15:25:15 +0000 (15:25 +0000)]
Can't reproduce the server has gone away messages, Andrey. Please, have a look. I tried MySQL 5.1, 5.5 and 5.6.
Johannes Schlüter [Thu, 1 Sep 2011 14:52:55 +0000 (14:52 +0000)]
Improve speed of finding tests (no more CVS, we're on svn now)
Ulf Wendel [Thu, 1 Sep 2011 14:04:52 +0000 (14:04 +0000)]
Fix test. First element within the parantheses is a SQL status code. Should please gcov
Christian Stocker [Thu, 1 Sep 2011 13:43:38 +0000 (13:43 +0000)]
added a NEWS entry about the change in XSL
Ulf Wendel [Thu, 1 Sep 2011 13:37:49 +0000 (13:37 +0000)]
Fixing tests debug output
Pierre Joye [Thu, 1 Sep 2011 13:28:00 +0000 (13:28 +0000)]
- use %s, bad c/p
Ulf Wendel [Thu, 1 Sep 2011 13:17:16 +0000 (13:17 +0000)]
Check if MySQL server supports SSL and, if using mysqlnd, check if PHP streams will support SSL
Hannes Magnusson [Thu, 1 Sep 2011 13:06:12 +0000 (13:06 +0000)]
Move the XFAILED printout way up the list so we can actually notice the really failing ones
Pierre Joye [Thu, 1 Sep 2011 12:49:26 +0000 (12:49 +0000)]
- fix test
Pierre Joye [Thu, 1 Sep 2011 12:42:54 +0000 (12:42 +0000)]
- win and nux work the same way
Pierre Joye [Thu, 1 Sep 2011 12:35:41 +0000 (12:35 +0000)]
- force LF eol style as this test depends on it
Johannes Schlüter [Thu, 1 Sep 2011 12:04:54 +0000 (12:04 +0000)]
- Fix TSRM build
Ulf Wendel [Thu, 1 Sep 2011 11:48:18 +0000 (11:48 +0000)]
Simplify mysqlnd debug trace marginally for debugging
Pierre Joye [Thu, 1 Sep 2011 11:40:37 +0000 (11:40 +0000)]
- consider that bison 2.x is used. updated our ci boxes accordingly (windows too), which were using 1.4
Pierre Joye [Thu, 1 Sep 2011 11:40:23 +0000 (11:40 +0000)]
- consider that bison 2.x is used. updated our ci boxes accordingly (windows too), which were using 1.4
Ulf Wendel [Thu, 1 Sep 2011 11:27:14 +0000 (11:27 +0000)]
Hack around http://bugs.mysql.com/62320
Hannes Magnusson [Thu, 1 Sep 2011 11:00:51 +0000 (11:00 +0000)]
Seeing thousands of PASS tests flying by is meaningless.
Add an option to only print certain result groups.
Pierre Joye [Thu, 1 Sep 2011 09:33:43 +0000 (09:33 +0000)]
- force LF eol style as this test depends on it
Matteo Beccati [Thu, 1 Sep 2011 00:45:56 +0000 (00:45 +0000)]
Fixed test randomly failing due to different error messages. Also added a skip
section in case "fakeURL" resolves to an IP address, as some empty error
messages and HTML can be seen in the reports
Johannes Schlüter [Wed, 31 Aug 2011 23:42:04 +0000 (23:42 +0000)]
- Change PDO_oci connection defaults to match Oracle XE defaults, higher chance
to work out of the box
Pierre Joye [Wed, 31 Aug 2011 23:12:35 +0000 (23:12 +0000)]
- error msg may vary (parser)
Matteo Beccati [Wed, 31 Aug 2011 22:38:20 +0000 (22:38 +0000)]
Fixed test. Apparently it was already fixed on PHP_5_3 98% of the cases and
the fix wasn't ported to trunk/PHP_5_4. However the very same test was failing
on my FreeBSD 6.2 box, that appears to also decode the "ks_c_5601-1987" parts,
so I've made the test more generic.
Matteo Beccati [Wed, 31 Aug 2011 22:12:53 +0000 (22:12 +0000)]
Fixed false positive failure on gcov
Johannes Schlüter [Wed, 31 Aug 2011 20:30:08 +0000 (20:30 +0000)]
- Register extensions using mysqlnd (mysql, myslqi, pdo_mysql) with mysqlnd
Boris Lytochkin [Wed, 31 Aug 2011 20:16:56 +0000 (20:16 +0000)]
tune text align
Pierre Joye [Wed, 31 Aug 2011 20:16:11 +0000 (20:16 +0000)]
- error msg may vary (parser)
Pierre Joye [Wed, 31 Aug 2011 20:14:13 +0000 (20:14 +0000)]
- error msg may vary (parser)
Pierre Joye [Wed, 31 Aug 2011 20:10:22 +0000 (20:10 +0000)]
- error msg may vary
Pierre Joye [Wed, 31 Aug 2011 20:07:44 +0000 (20:07 +0000)]
- typo
Boris Lytochkin [Wed, 31 Aug 2011 19:45:45 +0000 (19:45 +0000)]
test coverage for SNMP::ERRNO_MULTIPLE_SET_QUERIES
Tjerk Anne Meesters [Wed, 31 Aug 2011 18:26:23 +0000 (18:26 +0000)]
Fixed expect format to ignore the textual error message from readlink().
Bug #55555
Andrey Hristov [Wed, 31 Aug 2011 18:18:23 +0000 (18:18 +0000)]
Fix valgrind error (sending of initialized bytes over the network).
When the compression was successful the compressed data + uninitialized data
at the end was sent to the server, because the length of the compressed payload
wasn't correctly calculated (actually the length of the uncompressed payload as assumed).
However, the uncompress() function has internally the length of the real payload and skips
the binary trash at the end - thus no data damage occurs!
Johannes Schlüter [Wed, 31 Aug 2011 15:46:07 +0000 (15:46 +0000)]
- Update PDO_mysql error mapping
Pierre Joye [Wed, 31 Aug 2011 15:44:01 +0000 (15:44 +0000)]
- force LF eol style as this test depends on it
Pierre Joye [Wed, 31 Aug 2011 15:42:42 +0000 (15:42 +0000)]
- force LF eol style as this test depends on it
Johannes Schlüter [Wed, 31 Aug 2011 15:25:18 +0000 (15:25 +0000)]
- Be less restrictive on the SQL state, some server versions might fool us
Ulf Wendel [Wed, 31 Aug 2011 13:59:28 +0000 (13:59 +0000)]
Please, whenever changing the mysqli API update the tests. Don't wait for me to catch up.
Ulf Wendel [Wed, 31 Aug 2011 13:50:58 +0000 (13:50 +0000)]
New utility function to check for '[1148] The used command is not allowed with this MySQL version' during SKIPIF
Rui Hirokawa [Wed, 31 Aug 2011 13:19:03 +0000 (13:19 +0000)]
MFH: added alias for *-mobile encodings.
Ulf Wendel [Wed, 31 Aug 2011 12:58:25 +0000 (12:58 +0000)]
Check if server supports LOAD DATA LOCAL INFILE to catch [1148] The used command is not allowed with this MySQL version during SKIPIF
Pierre Joye [Wed, 31 Aug 2011 12:27:11 +0000 (12:27 +0000)]
- force LF eol style as this test depends on it (size)
Christian Stocker [Wed, 31 Aug 2011 11:44:22 +0000 (11:44 +0000)]
simplexml->query returns empty array if no nodes were found
and false if libxml thinks the xpath-expression was invalid.
Behaves now the same like DomXPath and fixes Bug #48601
Adjusted a test to reflect that change
Pierre Joye [Wed, 31 Aug 2011 11:26:27 +0000 (11:26 +0000)]
- force LF eol style as this test depends on it (size)
Pierre Joye [Wed, 31 Aug 2011 11:25:09 +0000 (11:25 +0000)]
- force LF eol style as this test depends on it (size)
Pierre Joye [Wed, 31 Aug 2011 11:18:13 +0000 (11:18 +0000)]
- parser version independent
Pierre Joye [Wed, 31 Aug 2011 11:16:25 +0000 (11:16 +0000)]
- parser version independent
Johannes Schlüter [Wed, 31 Aug 2011 11:11:08 +0000 (11:11 +0000)]
- Fix Bug #55550 mysql.trace_mode miscounts result sets,
fixes ext/mysql/tests/mysql_deprecated_api.phpt
Pierre Joye [Wed, 31 Aug 2011 10:46:20 +0000 (10:46 +0000)]
- use only the test name
Pierre Joye [Wed, 31 Aug 2011 10:43:39 +0000 (10:43 +0000)]
- use only the test name
Pierre Joye [Wed, 31 Aug 2011 10:30:39 +0000 (10:30 +0000)]
- use only the test name
Pierre Joye [Wed, 31 Aug 2011 10:22:18 +0000 (10:22 +0000)]
- use only test name
Hannes Magnusson [Wed, 31 Aug 2011 10:09:54 +0000 (10:09 +0000)]
Remove annoying extra newline on XFAIL summaries
and sync text to trunk
Pierre Joye [Wed, 31 Aug 2011 10:08:55 +0000 (10:08 +0000)]
- force LF eol style as this test depends on it (size)
Pierre Joye [Wed, 31 Aug 2011 10:05:54 +0000 (10:05 +0000)]
- make it parser version independent
Christian Stocker [Wed, 31 Aug 2011 09:14:08 +0000 (09:14 +0000)]
- Set an empty string and avoid warnings and stopping of the transformation process
when a php function can't be called.
- Fixed tests.
Christian Stocker [Wed, 31 Aug 2011 08:58:12 +0000 (08:58 +0000)]
fixed libxml tests
Boris Lytochkin [Wed, 31 Aug 2011 08:36:22 +0000 (08:36 +0000)]
more tuning based on discussion in FR #40816:
* parse all OIDs earlier, detect all wrong OIDs before any query
is made (GET-operations)
* introduce ERRNO_MULTIPLE_SET_QUERIES:
warn if request contains more OIDs than max_oids and SET operation
(and type&value checks) will be done in chunks.
fix set method when request contains more OIDs than max_oids (2nd and
subsequent chunk were ignored)
Pierre Joye [Wed, 31 Aug 2011 08:35:57 +0000 (08:35 +0000)]
- force LF eol style as this test depends on it (size)
Rasmus Lerdorf [Wed, 31 Aug 2011 07:43:58 +0000 (07:43 +0000)]
Since we are creating this csv file locally this should obviously
be a a LOCAL infile here for this to work reliably. For robustness
this should probably have a check for the local-infile server
setting as well.
Stanislav Malyshev [Wed, 31 Aug 2011 07:34:07 +0000 (07:34 +0000)]
fix strptime tests for mac
Rasmus Lerdorf [Wed, 31 Aug 2011 07:19:20 +0000 (07:19 +0000)]
Fix test. I have no idea why the test author didn't expect to get back
latin2 when the test is setting it to latin2 here. Both PHP and MySQL
are doing the right thing here.
Stanislav Malyshev [Wed, 31 Aug 2011 07:10:17 +0000 (07:10 +0000)]
test fixes on Darwin
Gustavo André dos Santos Lopes [Wed, 31 Aug 2011 06:32:05 +0000 (06:32 +0000)]
- Fixed test htmlentities15.phpt.