]> granicus.if.org Git - php/log
php
19 years agoremember ? -> :pdox mapping so that binds by position can be mapped to names if required.
Wez Furlong [Tue, 12 Jul 2005 03:19:44 +0000 (03:19 +0000)]
remember ? -> :pdox mapping so that binds by position can be mapped to names if required.

19 years agofix bug when query was rewritten from ? to :name
Wez Furlong [Tue, 12 Jul 2005 03:18:17 +0000 (03:18 +0000)]
fix bug when query was rewritten from ? to :name

19 years agoimprove handling of bound input parameters when no maximal length value is set;
Wez Furlong [Tue, 12 Jul 2005 02:43:39 +0000 (02:43 +0000)]
improve handling of bound input parameters when no maximal length value is set;
default to 4000 as the maximal length, which is the biggest size possible
without using a LONG type (if you specify anything larger than this, you'll end
up with ORA-1461).

Don't assume that all parameters were output parameters after execution, as
this would clobber the input values when used in a loop.

19 years agoexpand oracle null handling compatability by offering the ability to convert
Wez Furlong [Tue, 12 Jul 2005 02:40:59 +0000 (02:40 +0000)]
expand oracle null handling compatability by offering the ability to convert
NULLs into empty strings as well as the other way around.  It still doesn't
help a great deal in the long run, but at least the option is there.

Make sure hash tables are nulled out to avoid double freeing them.

19 years agoshould always bind columns after execute.
Wez Furlong [Tue, 12 Jul 2005 02:38:39 +0000 (02:38 +0000)]
should always bind columns after execute.
There're probably more of these to fix.

19 years agoMake references misuse emit E_NOTICE rather E_STRICT to be compatible with
Ilia Alshanetsky [Mon, 11 Jul 2005 18:44:37 +0000 (18:44 +0000)]
Make references misuse emit E_NOTICE rather E_STRICT to be compatible with
PHP 4.4.0

19 years agoA bit of further tuning.
Ilia Alshanetsky [Mon, 11 Jul 2005 14:46:01 +0000 (14:46 +0000)]
A bit of further tuning.

19 years agofix a segfault with the following script:
Andrey Hristov [Mon, 11 Jul 2005 14:14:56 +0000 (14:14 +0000)]
fix a segfault with the following script:
<?php
$dbh = new PDO('mysql:dbname=test;host=localhost', "root", "secret");
$what = 1;
$stmt = $dbh->prepare('select a, b, c from t123 where a=:what');
$stmt->bindParam(1, $what, PDO_PARAM_INT, 12);
var_dump($stmt->execute());
var_dump($stmt->fetchObject());
?>

19 years ago- mysql_affected_rows() already returns an unsigned long
Andrey Hristov [Mon, 11 Jul 2005 13:52:13 +0000 (13:52 +0000)]
- mysql_affected_rows() already returns an unsigned long
- optimize a bit the real_escape
- handle the situation when libmysql bails out because of lack of memory

19 years agoRemoving the automake-related output. The build no longer uses automake.
Jon Parise [Mon, 11 Jul 2005 06:32:02 +0000 (06:32 +0000)]
Removing the automake-related output.  The build no longer uses automake.

While I'm here, rewrite the PHP_AUTOCONF-related code to more closely
match the conventions used elsewhere in this section.  The result should
be functionally the same as before.

19 years agoChangeLog update
<changelog@php.net> [Mon, 11 Jul 2005 05:39:30 +0000 (05:39 +0000)]
ChangeLog update

19 years agoCompiler warning fix.
Ilia Alshanetsky [Mon, 11 Jul 2005 03:24:28 +0000 (03:24 +0000)]
Compiler warning fix.

19 years ago- Moved ext/dbx and ext/ircg to PECL.
foobar [Sun, 10 Jul 2005 19:05:09 +0000 (19:05 +0000)]
- Moved ext/dbx and ext/ircg to PECL.

19 years agotouch
foobar [Sun, 10 Jul 2005 17:56:35 +0000 (17:56 +0000)]
touch

19 years agoFix #33624, mysterious crashes on shutdown on win32
Wez Furlong [Sun, 10 Jul 2005 14:56:36 +0000 (14:56 +0000)]
Fix #33624, mysterious crashes on shutdown on win32

19 years ago- use three digit octal character definition
Marcus Boerger [Sun, 10 Jul 2005 14:04:20 +0000 (14:04 +0000)]
- use three digit octal character definition

19 years agoChangeLog update
<changelog@php.net> [Sun, 10 Jul 2005 05:39:03 +0000 (05:39 +0000)]
ChangeLog update

19 years agostmt->column_count is set only if the result has been already bound, don't
Andrey Hristov [Sun, 10 Jul 2005 02:58:51 +0000 (02:58 +0000)]
stmt->column_count is set only if the result has been already bound, don't
rely on that and use the real count which is kept in st_mysql_stmt
#of course using internal values is bad idea but the comment already says
#that this should be fixed

19 years agook, calculate max_length only in case of a blob (normal, medium, long).
Andrey Hristov [Sun, 10 Jul 2005 02:49:14 +0000 (02:49 +0000)]
ok, calculate max_length only in case of a blob (normal, medium, long).
in case of a varchar lets allocate 255 bytes and be quick without asking
libmysql to update max_length

19 years agoif the result set is buffered ask libmysql to compute the lengths, so
Andrey Hristov [Sun, 10 Jul 2005 02:20:26 +0000 (02:20 +0000)]
if the result set is buffered ask libmysql to compute the lengths, so
later allocate buffer as big as the largest value of the column in the
result set (max_legth positive).

19 years agoGet precise lengths from the server, as suggested by Andrey.
Wez Furlong [Sun, 10 Jul 2005 02:00:35 +0000 (02:00 +0000)]
Get precise lengths from the server, as suggested by Andrey.

19 years agoRelated to #33624. Crashes for me on shutdown, but seems ok for the rest of the...
Wez Furlong [Sat, 9 Jul 2005 18:52:36 +0000 (18:52 +0000)]
Related to #33624.  Crashes for me on shutdown, but seems ok for the rest of the world.

19 years agoChangeLog update
<changelog@php.net> [Sat, 9 Jul 2005 05:37:17 +0000 (05:37 +0000)]
ChangeLog update

19 years agoAdded safety checks.
Ilia Alshanetsky [Sat, 9 Jul 2005 05:08:54 +0000 (05:08 +0000)]
Added safety checks.

19 years agoMake cursor closer work with emulation as well.
Ilia Alshanetsky [Sat, 9 Jul 2005 05:04:43 +0000 (05:04 +0000)]
Make cursor closer work with emulation as well.

19 years agodon't free the cols until we know we have another rowset
Wez Furlong [Sat, 9 Jul 2005 05:01:14 +0000 (05:01 +0000)]
don't free the cols until we know we have another rowset

19 years agoAdded cursor closer handler.
Ilia Alshanetsky [Sat, 9 Jul 2005 04:54:04 +0000 (04:54 +0000)]
Added cursor closer handler.
Fixed memory leak.

19 years agoAdded missing header check.
Ilia Alshanetsky [Sat, 9 Jul 2005 04:43:16 +0000 (04:43 +0000)]
Added missing header check.

19 years agocursor_closer for SQLite.
Ilia Alshanetsky [Sat, 9 Jul 2005 04:37:27 +0000 (04:37 +0000)]
cursor_closer for SQLite.

19 years agoImplement cursor_closer for MySQL driver.
Ilia Alshanetsky [Sat, 9 Jul 2005 04:30:49 +0000 (04:30 +0000)]
Implement cursor_closer for MySQL driver.

19 years agofor the transactions test case, detect working transactions before attempting to...
Wez Furlong [Sat, 9 Jul 2005 04:28:45 +0000 (04:28 +0000)]
for the transactions test case, detect working transactions before attempting to run the tests.
Additional ugliness required because mysql does stupid stuff like this:

mysql> CREATE TABLE foo (id int) TYPE=innodb;
Query OK, 0 rows affected, 2 warnings (0.00 sec)
mysql> SHOW CREATE TABLE foo;
CREATE TABLE `foo` (
  `id` int(11) default NULL
) ENGINE=MyISAM DEFAULT CHARSET=latin1

In addition, BEGIN, COMMIT and ROLLBACK all succeed, even when no tables
support transactions.

19 years agoFixed memory leak
Ilia Alshanetsky [Sat, 9 Jul 2005 04:21:14 +0000 (04:21 +0000)]
Fixed memory leak

19 years agoFix two bugs:
Wez Furlong [Sat, 9 Jul 2005 03:54:13 +0000 (03:54 +0000)]
Fix two bugs:

- execute() would not re-fetch meta data after nextRowset() had been called.

- buffered mode would only be enabled on the first execute on a statement handle.

19 years agoAdded:
Wez Furlong [Sat, 9 Jul 2005 03:52:41 +0000 (03:52 +0000)]
Added:

proto bool PDOStatement::closeCursor()
Closes the cursor, leaving the statement ready for re-execution.

The purpose of the function is to free up the connection to the server so that
other queries may be issued, but leaving the statement in a state that it can
be re-executed.

This is implemented either as an optional driver specific method (allowing for
maximum efficiency), or as the generic PDO fallback if no driver specific
function is installed.

The PDO generic fallback is semantically the same as writing the following code
in your PHP script:

do {
while ($stmt->fetch())
;
if (!$stmt->nextRowset())
break;
} while (true);

19 years agoSafer way of allowing 'blank' user/pass.
Ilia Alshanetsky [Sat, 9 Jul 2005 02:53:07 +0000 (02:53 +0000)]
Safer way of allowing 'blank' user/pass.

# Thanks Wez.

19 years agoshow ugly full path for failed redirected tests
foobar [Sat, 9 Jul 2005 02:41:19 +0000 (02:41 +0000)]
show ugly full path for failed redirected tests

19 years ago- No need to use convert_to_string_ex() when dealing with strings
foobar [Sat, 9 Jul 2005 00:59:49 +0000 (00:59 +0000)]
- No need to use convert_to_string_ex() when dealing with strings

19 years ago- Fixed bug #33588 (LDAP: RootDSE query not possible).
foobar [Sat, 9 Jul 2005 00:46:45 +0000 (00:46 +0000)]
- Fixed bug #33588 (LDAP: RootDSE query not possible).

19 years agoFix needed for PostgreSQL driver to allow $user & $pass to be supplied via
Ilia Alshanetsky [Fri, 8 Jul 2005 23:33:35 +0000 (23:33 +0000)]
Fix needed for PostgreSQL driver to allow $user & $pass to be supplied via
DSN as NULL.

19 years agoSanitazione
foobar [Fri, 8 Jul 2005 21:46:23 +0000 (21:46 +0000)]
Sanitazione

19 years agofix leak
Wez Furlong [Fri, 8 Jul 2005 20:45:19 +0000 (20:45 +0000)]
fix leak

19 years ago- Regenerate using re2c 0.9.8
Marcus Boerger [Fri, 8 Jul 2005 20:38:13 +0000 (20:38 +0000)]
- Regenerate using re2c 0.9.8

19 years ago- std not needed -> avoid warning
Marcus Boerger [Fri, 8 Jul 2005 20:35:41 +0000 (20:35 +0000)]
- std not needed -> avoid warning

19 years ago- Fix *printf(%lld) under windows
Marcus Boerger [Fri, 8 Jul 2005 20:19:38 +0000 (20:19 +0000)]
- Fix *printf(%lld) under windows

19 years agoFix Bug #33111
Wez Furlong [Fri, 8 Jul 2005 19:28:52 +0000 (19:28 +0000)]
Fix Bug #33111

19 years agoFix #33618
Wez Furlong [Fri, 8 Jul 2005 19:25:37 +0000 (19:25 +0000)]
Fix #33618

19 years agoFix PECL #4753
Wez Furlong [Fri, 8 Jul 2005 19:05:57 +0000 (19:05 +0000)]
Fix PECL #4753

19 years agofix tsrm build
Antony Dovgal [Fri, 8 Jul 2005 18:16:46 +0000 (18:16 +0000)]
fix tsrm build

19 years agorewrite original names to our preferred format
Wez Furlong [Fri, 8 Jul 2005 17:01:20 +0000 (17:01 +0000)]
rewrite original names to our preferred format

19 years agogenerated file
Wez Furlong [Fri, 8 Jul 2005 17:00:48 +0000 (17:00 +0000)]
generated file

19 years agoFor named-parameter-to-named-parameter rewrites, we need to map the original
Wez Furlong [Fri, 8 Jul 2005 17:00:28 +0000 (17:00 +0000)]
For named-parameter-to-named-parameter rewrites, we need to map the original
names to the new names.

19 years agoSome of us don't have PostgreSQL 8.0 :)
Ilia Alshanetsky [Fri, 8 Jul 2005 16:20:13 +0000 (16:20 +0000)]
Some of us don't have PostgreSQL 8.0 :)

19 years agoMissing bit of the previous patch.
Ilia Alshanetsky [Fri, 8 Jul 2005 16:17:04 +0000 (16:17 +0000)]
Missing bit of the previous patch.

19 years agoFixed compiler warning.
Ilia Alshanetsky [Fri, 8 Jul 2005 16:06:05 +0000 (16:06 +0000)]
Fixed compiler warning.

19 years agoAdd early support for native prepared statements in pgsql.
Wez Furlong [Fri, 8 Jul 2005 15:27:34 +0000 (15:27 +0000)]
Add early support for native prepared statements in pgsql.
Note that some tests now fail; if we can't resolve this in time for the beta,
the prepare code should be disabled (I'll add a flag for this later today).

19 years agoFix a shutdown order issue I saw in the pgsql driver. Hope this doesn't mess
Wez Furlong [Fri, 8 Jul 2005 15:25:15 +0000 (15:25 +0000)]
Fix a shutdown order issue I saw in the pgsql driver.  Hope this doesn't mess
up something in the OCI driver; I think I've been here before.

19 years agoAnd the generated version of the parser
Wez Furlong [Fri, 8 Jul 2005 15:24:42 +0000 (15:24 +0000)]
And the generated version of the parser

19 years agoadd a bit of a hack to cater for pgsql prepared statements.
Wez Furlong [Fri, 8 Jul 2005 15:24:21 +0000 (15:24 +0000)]
add a bit of a hack to cater for pgsql prepared statements.

These are effectively named statements with strong constraints on the naming
format.  We cater for this in a fairly generic way: allow a driver to replace
the format string we use to generate names from positional parameters.  In
addition, if that format is set, we always force a rewrite from regular names
to the strongly enforced names.

19 years agoand as a result, need to update the expected data
Wez Furlong [Fri, 8 Jul 2005 15:21:43 +0000 (15:21 +0000)]
and as a result, need to update the expected data

19 years agowe declare the column as NOT NULL, so inserting NULL is not a good idea.
Wez Furlong [Fri, 8 Jul 2005 15:20:18 +0000 (15:20 +0000)]
we declare the column as NOT NULL, so inserting NULL is not a good idea.

19 years agomake use of T token
Antony Dovgal [Fri, 8 Jul 2005 12:39:51 +0000 (12:39 +0000)]
make use of T token

19 years agofix #33597 (setcookie() "expires" date format doesn't comply with RFC)
Antony Dovgal [Fri, 8 Jul 2005 12:30:24 +0000 (12:30 +0000)]
fix #33597 (setcookie() "expires" date format doesn't comply with RFC)

19 years agorename php_format_date() to date_format()
Antony Dovgal [Fri, 8 Jul 2005 12:26:30 +0000 (12:26 +0000)]
rename php_format_date() to date_format()
add wrapper for date_format() to use it in other extensions

19 years agofix #33614 (memory leak in new strftime())
Antony Dovgal [Fri, 8 Jul 2005 10:23:33 +0000 (10:23 +0000)]
fix #33614 (memory leak in new strftime())

19 years agoFixed HTTP basic authentication headers during subrequsts to xsd files
Dmitry Stogov [Fri, 8 Jul 2005 09:36:28 +0000 (09:36 +0000)]
Fixed HTTP basic authentication headers during subrequsts to xsd files

19 years agoMake sure we get a WARNING and not an ERROR here :)
foobar [Fri, 8 Jul 2005 08:53:25 +0000 (08:53 +0000)]
Make sure we get a WARNING and not an ERROR here :)

19 years agofix test when session.save_handler is "user"
foobar [Fri, 8 Jul 2005 07:57:47 +0000 (07:57 +0000)]
fix test when session.save_handler is "user"

19 years agofix test when session.save_handler is "user"
foobar [Fri, 8 Jul 2005 07:48:57 +0000 (07:48 +0000)]
fix test when session.save_handler is "user"

19 years agoFixed support for <any> in base type of extension
Dmitry Stogov [Fri, 8 Jul 2005 07:16:56 +0000 (07:16 +0000)]
Fixed support for <any> in base type of extension

19 years agoChangeLog update
<changelog@php.net> [Fri, 8 Jul 2005 05:38:53 +0000 (05:38 +0000)]
ChangeLog update

19 years agoAdd a PDO_ATTR_STRINGIFY_FETCHES attribute, which is used to convert integer or
Wez Furlong [Fri, 8 Jul 2005 04:13:00 +0000 (04:13 +0000)]
Add a PDO_ATTR_STRINGIFY_FETCHES attribute, which is used to convert integer or
floating point values into strings during fetch.  This is a compatibility hack
for drivers that return native types rather than string representations.

We use this flag in the test suite to persuade postgres tests to pass.

19 years agoAdded pg_fetch_all_columns() function to fetch all values of a column from
Ilia Alshanetsky [Fri, 8 Jul 2005 00:40:32 +0000 (00:40 +0000)]
Added pg_fetch_all_columns() function to fetch all values of a column from
a result cursor.

19 years ago- Fixed stupid bug in PHP_INSTALL_HEADERS() macro: "foo bar" != foo bar
foobar [Thu, 7 Jul 2005 23:18:22 +0000 (23:18 +0000)]
- Fixed stupid bug in PHP_INSTALL_HEADERS() macro: "foo bar" != foo bar
- ext/date/lib/timelib_config.h is also installed now

19 years agodo not set error_code to HY000 when there was NO error
Antony Dovgal [Thu, 7 Jul 2005 23:03:22 +0000 (23:03 +0000)]
do not set error_code to HY000 when there was NO error
reset both handle & statement error codes to 0

19 years agofree parameter when requested to do so
Antony Dovgal [Thu, 7 Jul 2005 23:02:22 +0000 (23:02 +0000)]
free parameter when requested to do so

19 years ago1st set of fixes to make tests work with MySQL
Ilia Alshanetsky [Thu, 7 Jul 2005 18:49:58 +0000 (18:49 +0000)]
1st set of fixes to make tests work with MySQL

# Additional changes pending

19 years agoMake prepared statements obey buffering flag in MySQL.
Ilia Alshanetsky [Thu, 7 Jul 2005 17:53:34 +0000 (17:53 +0000)]
Make prepared statements obey buffering flag in MySQL.

19 years agoFixed SIGSEGV on decoding <any> elements
Dmitry Stogov [Thu, 7 Jul 2005 17:29:25 +0000 (17:29 +0000)]
Fixed SIGSEGV on decoding <any> elements

19 years agopopulate the new dbh->driver field.
Wez Furlong [Thu, 7 Jul 2005 17:08:01 +0000 (17:08 +0000)]
populate the new dbh->driver field.
Implement PDO_ATTR_DRIVER_NAME.

19 years agoFixed memory leak
Ilia Alshanetsky [Thu, 7 Jul 2005 16:27:46 +0000 (16:27 +0000)]
Fixed memory leak

19 years agoadd an attribute for returning the driver name
Wez Furlong [Thu, 7 Jul 2005 16:24:51 +0000 (16:24 +0000)]
add an attribute for returning the driver name

19 years agoFixed bug #33512 (Add missing support for isset()/unset() overloading to complement...
Dmitry Stogov [Thu, 7 Jul 2005 16:09:55 +0000 (16:09 +0000)]
Fixed bug #33512 (Add missing support for isset()/unset() overloading to complement the property get/set methods). Now empty($obj[...]) works proper but in addition it may call offsetGet() method.

19 years agoFixed bug #33512 (Add missing support for isset()/unset() overloading to complement...
Dmitry Stogov [Thu, 7 Jul 2005 16:07:09 +0000 (16:07 +0000)]
Fixed bug #33512 (Add missing support for isset()/unset() overloading to complement the property get/set methods)

19 years agoFixed memory leak.
Ilia Alshanetsky [Thu, 7 Jul 2005 15:54:00 +0000 (15:54 +0000)]
Fixed memory leak.

19 years agozend_stream_getc uses fread internally. NetWare LibC fread reads 4(Which I believe...
Anantha Kesari H Y [Thu, 7 Jul 2005 15:43:50 +0000 (15:43 +0000)]
zend_stream_getc uses fread internally. NetWare LibC fread reads 4(Which I believe EOT) for EOF(^D) character. This happens when fread is asked to read one and only character as is the case with cl interactive mode.
-- Kamesh

19 years agoremove old test files
Wez Furlong [Thu, 7 Jul 2005 15:36:00 +0000 (15:36 +0000)]
remove old test files

19 years agousing new tests
Wez Furlong [Thu, 7 Jul 2005 15:34:46 +0000 (15:34 +0000)]
using new tests

19 years agouse new tests
Wez Furlong [Thu, 7 Jul 2005 15:33:06 +0000 (15:33 +0000)]
use new tests

19 years agoStop on the 1st skip.
Ilia Alshanetsky [Thu, 7 Jul 2005 15:32:32 +0000 (15:32 +0000)]
Stop on the 1st skip.

19 years agoadopt new tests
Wez Furlong [Thu, 7 Jul 2005 15:20:06 +0000 (15:20 +0000)]
adopt new tests

19 years agofix #33605 (substr_compare() crashes with negative offset & length)
Antony Dovgal [Thu, 7 Jul 2005 15:19:40 +0000 (15:19 +0000)]
fix #33605 (substr_compare() crashes with negative offset & length)

19 years agoFixed bug #28072 (static array with some constant keys will be incorrectly ordered).
Dmitry Stogov [Thu, 7 Jul 2005 15:16:57 +0000 (15:16 +0000)]
Fixed bug #28072 (static array with some constant keys will be incorrectly ordered).

19 years agoskip if we can't connect (include reason in skip output)
Wez Furlong [Thu, 7 Jul 2005 15:15:01 +0000 (15:15 +0000)]
skip if we can't connect (include reason in skip output)

19 years agoReturn an empty array rather then FALSE in fetchAll() on no results.
Ilia Alshanetsky [Thu, 7 Jul 2005 15:14:10 +0000 (15:14 +0000)]
Return an empty array rather then FALSE in fetchAll() on no results.

19 years agoFixed memory leak, after SQLITE_DONE sqlite3_data_count() always returns 0.
Ilia Alshanetsky [Thu, 7 Jul 2005 15:13:28 +0000 (15:13 +0000)]
Fixed memory leak, after SQLITE_DONE sqlite3_data_count() always returns 0.

19 years agoadopt new tests
Wez Furlong [Thu, 7 Jul 2005 15:04:45 +0000 (15:04 +0000)]
adopt new tests

19 years agoDB2 wants us to use the proper ODBC 3 function for closing a cursor.
Wez Furlong [Thu, 7 Jul 2005 14:20:04 +0000 (14:20 +0000)]
DB2 wants us to use the proper ODBC 3 function for closing a cursor.

19 years agoNeed to state NOT NULL here, as DB2 won't allow the UNIQUE constraint without it.
Wez Furlong [Thu, 7 Jul 2005 14:09:58 +0000 (14:09 +0000)]
Need to state NOT NULL here, as DB2 won't allow the UNIQUE constraint without it.

19 years agomake column names lower case
Wez Furlong [Thu, 7 Jul 2005 13:59:34 +0000 (13:59 +0000)]
make column names lower case

19 years agoadopt common tests
Wez Furlong [Thu, 7 Jul 2005 13:58:35 +0000 (13:58 +0000)]
adopt common tests