This issue is created automatically through a Github pull request on behalf of adrienbrault:

Url: https://github.com/doctrine/dbal/pull/125

Message:

Hey guys,

I started to fix a class type inside Connection and then i went ahead to see what i could also improve!

Adrien

Can not set charset options for table creation.
Now I am fix thix problem in patch.
File: doctrine-dbal/lib/Doctrine/DBAL/Schema/Schema.php

#!/usr/bin/php
<?php
/**

• Doctrine DBAL Patch
• fix table creation query encode
• (c) _Nicolay
  */
  system ('cp ./vendor/doctrine-dbal/lib/Doctrine/DBAL/Schema/Schema.php ./vendor/doctrine-dbal/lib/Doctrine/DBAL/Schema/Schema.php.Orig');

$data = file_get_contents('./vendor/doctrine-dbal/lib/Doctrine/DBAL/Schema/Schema.php');
$data = str_replace("new Table(\$tableName);",
"new Table(\$tableName,array(),array(),array(),0,array('charset'=>'utf8','collate'=>'utf8_general_ci'));",
$data);
file_put_contents('./vendor/doctrine-dbal/lib/Doctrine/DBAL/Schema/Schema.php', $data);

Hi Guys,

I was working with symfony command line and created the database with the app/console doctrine:create:database

Though the characterset I specified was UTF-8 , it created Latin character set .

@elliot was right too, the create database is not using any character set https://github.com/doctrine/dbal/blob/master/lib/Doctrine/DBAL/Platforms/MySqlPlatform.php#L328

You can see the issue created at https://github.com/symfony/symfony/issues/2044 and later Fabien confirmed the bug is related to Doctrine .

I am not sure its a Bug or a feature I am asking .

Thanks

In my symfony 2.1 project i was including JMSPaymentCoreBundle and did the vanilla installation.

After that i ran

php app/console doctrine:schema:update --dump-sql
php app/console doctrine:schema:update --force

A lot of schema updates were done. I ran it again, and two schema updates were still showing up.
Running the command again, and they still show up.

I then tried to run the two schema updates against my database and MySQL complains with the following error:

Error: 1101 - BLOB/TEXT column 'extended_data' can't have a default value

According to the Documentation BLOB/TEXT indeed cannot have default values. I am not sure why Doctrine thinks it has to set this default value.

The statements look like this:

sascha@debian:/var/www/myproject/Symfony$ php app/console doctrine:schema:update --dump-sql
ALTER TABLE payment_instructions CHANGE extended_data extended_data LONGTEXT NOT NULL COMMENT '(DC2Type:extended_payment_data)';
ALTER TABLE financial_transactions CHANGE extended_data extended_data LONGTEXT DEFAULT NULL COMMENT '(DC2Type:extended_payment_data)'

I already reported this problem in the JMSPaymentCoreBundle Issue Queue where i then was referred to over here.

Also, i am not sure if Doctrine DBAL is the best match for this issue, so if it is wrong please move it to the right project issue queue.

Thank you very much and best of Regards!
Sascha

This issue is created automatically through a Github pull request on behalf of VladanStef:

Url: https://github.com/doctrine/dbal/pull/198

Message:

I think default table collation should be changed to `utf8_unicode_ci`, as it handles properly a much wider array of characters (Cyrillic, for example) at a generally negligible cost of performance.

`utf8_general_ci` simply does not work properly with too many languages and should be avoided as a default.

This issue is created automatically through a Github pull request on behalf of naitsirch:

Url: https://github.com/doctrine/dbal/pull/197

Message:

Bugfix DDC-2013 - Default Session Value ".," for "NLS_NUMERIC_CHARACTERS". This is needed because Oracle uses other characters for decimal separation depending on the servers locale.

See http://www.doctrine-project.org/jira/browse/DDC-2013

has been fixed in 3f5fda98299acdcda9965e8e797cbcd0490f0dcc, however it's not merged into 2.3.

When I insert decimals into database I get the following error message:

ORA-01722: invalid number

Oracle wants decimals (e.g. NUMBER(5,2)) separated by comma instead of a dot. I think the reason is that I use the german version of Windows. I tried to modify the following method of \Doctrine\DBAL\Driver\OCI8\OCI8Statement::bindValue() just for testing of course:

    public function bindValue($param, $value, $type = null)
    {
        if (is_float($value)) {
            $value = str_replace('.', ',', (string) $value); // <--
        }
        return $this->bindParam($param, $value, $type);
    }

With this modification I do not get the error anymore.
It would be great if a solution could be found. The strange thing is, if I query "SELECT * from nls_database_parameters where PARAMETER='NLS_NUMERIC_CHARACTERS'" I get ".," which means that a dot is already used as decimal separator. Maybe Oracle preferes the Locale of the OS?!
Maybe you could provide a setting which defines what separator should be used for decimals.

I found some Links that might be interesting regarding this issue:

• http://docs.oracle.com/cd/E19455-01/806-0169/overview-9/index.html (Oracle Docs - Decimal and Thousands Separators)
• https://forums.oracle.com/forums/thread.jspa?threadID=35398
• http://docs.oracle.com/cd/B28359_01/olap.111/b28126/dml_options072.htm
• https://forums.oracle.com/forums/thread.jspa?threadID=899739
• http://symfony-blog.driebit.nl/2010/09/doctrine-and-oracle-under-windows/

i have this error in version 2.3.0-RC1

in DBAL/STatement.php on aound line 191 you will find:

public function setFetchMode($fetchMode, $arg2 = null, $arg3 = null)

the problem is that php throws me a Exception:
Fatal error: Uncaught exception 'PDOException' with message 'SQLSTATE[HY000]: General error: fetch mode doesn't allow any extra arguments'

the reason for the exception is that setFetchMode doesnt support null as second/third argument.

somebody also described it in the php issue tracker:
https://bugs.php.net/bug.php?id=54545

Forgive me to doubt, but I think there may be a bug in MasterSlaveConnection.

It's easier to understand what I'm saying by debugging and tracing the flow, but I'll illustrate it with gists.

First, here is a simple service method to create a user. It opens up a transaction, persists the user, commits and returns. On error, if there is an active transaction, rollback. Here is the gist:

https://gist.github.com/3547674

The "$conn->beginTransaction();" line is where we trace through (the remainder of the service method is now irrelevant). Looking into MasterSlaveConnection.php, we see the method tries to connect to the master connection (call this point ###):

https://gist.github.com/3547720

Now looking in the next gist, we see what happens when "$this->connect('master');" is called. At this point it's not that interesting, the internal "$this->_conn" property is set to "master".

https://gist.github.com/3547750

Now here lies the bug I believe. "parent::beginTransaction();" is called. When looking into this method, we see that another call is made to connect but this time without "master" as the argument (i.e. connect to slave). This call to connect is made before incrementing the transaction nesting level.

https://gist.github.com/3547808

Now, I won't do another gist for "MasterSlaveConnection::connect", but if you refer to the file at line 13 https://gist.github.com/3547750#file_master_slave_connection.php, you will see that it checks the transaction nesting level and if it is there, forces master. However, we don't increment the level until after the method returns, so the slave is used. Ultimately, this results in the internal "$this->_conn" property set to the "slave" connection which violates our original action at ### above where we said we want to connect to "master".

Am I missing something here? Here is a gist the is a basic attempt at fixing this one method. It simply copies the code from the parent method except does not connect twice. I believe the same would have to occur for all the other methods unless it can be fixed once at the "MasterSlaveConnection::connect" level.

https://gist.github.com/3547880

I've just fleshed out "beginTransaction", "commit" and "rollBack" in "MasterSlaveConnection" by basically copying and pasting the code from the parent class and for my failing use case, this fixes the issue. However, it did require updating "Connection" slightly so that I had access to some private variables.

This issue is created automatically through a Github pull request on behalf of posulliv:

Url: https://github.com/doctrine/dbal/pull/194

Message:

When running the test suite against a clean PostgreSQL install on a 64-bit system I hit 2 issues:

• dropDatabase calls fail unless a database is manually created
• TypeConversionTest::testIdempotentDataConversion fails expecting bigint type to be converted to a string. PostgreSQL seems to detect 32 vs 64 bit systems and casts correctly for bigint types.

This small patch allows the test suite to run successfully for me against clean MySQL & PostgreSQL installs.

This issue is created automatically through a Github pull request on behalf of FabioBatSilva:

Url: https://github.com/doctrine/dbal/pull/190

Message:

Fix DDC-1978

http://www.doctrine-project.org/jira/browse/DDC-1978

This issue is created automatically through a Github pull request on behalf of stof:

Url: https://github.com/doctrine/dbal/pull/182

Message:

This adds a missing method in the Constraint interface. The method is already defined in AbstractAsset which is extended by both classes implementing the interface. the AbstractPlatform is calling this method on objects after checking they are implementing the interface. So either the method should be part of the interface, or all places using the interface as typehint should be refactored to check for the classes (examples can be found [here](https://github.com/doctrine/dbal/blob/master/lib/Doctrine/DBAL/Platforms/AbstractPlatform.php#L1029) and [there](https://github.com/doctrine/dbal/blob/master/lib/Doctrine/DBAL/Platforms/AbstractPlatform.php#L1254)).

Adding a method in an interface is technically a BC break but this method is not really a public extension point. what do you think @beberlei ?

This issue is created automatically through a Github pull request on behalf of stof:

Url: https://github.com/doctrine/dbal/pull/180

Message:

This fixes some phpdoc in the AbstractPlatform and in some child classes. I haven't checked all child classes to ensure that all phpdoc is right now, but the abstract one is now accurate (but a bunch of method are missing their phpdoc altogether)

• SQL Server/Azure for example requires strict differentiation between CLUSTERED/NON-CLUSTERED indexes.
• MySQL can create full text indexes
• etcetc.

SQL Azure has some platform specific SQL on top of SQL Server 2008 mainly to support Federations

Currently we only have support for PDO_SQLSRV, however some of the Azure/Advanced functionality strictly needs the sqlsrv extension.

My Symfony2 projects currently use main trunks of the Doctrine projects, the latest release merge into this branch was 2.2.0-BETA2, that's why I mentioned it in"Affected versions". Sorry if mistaken.

I would like to report that MSSQL Driver/Platform still does not support the 'uniquidentifier' field type of MSSQL.

Trying to introspect the database with
php app/console doctrine:mapping:import MMPTestDBBundle annotation
we get the error

[Doctrine\DBAL\DBALException]
  Unknown database type uniqueidentifier requested, Doctrine\DBAL\Platforms\SQLServer2008Platform may not support it.

There is a related issue, opened and fixed in Doctrine1 project
http://www.doctrine-project.org/jira/browse/DC-253

Going to fix this issue in the same way. Any help greatly appreciated.

Pull-Request was automatically synchronized: https://github.com/doctrine/dbal/pull/85

Drizzle platform support
Driver is using pdo_mysql

All DBAL tests complete except PortabilityTest.

This issue is created automatically through a Github pull request on behalf of dpb587:

Url: https://github.com/doctrine/dbal/pull/168

Message:

I noticed that DebugStack was adding an empty element when it was disabled. This PR changes it to match startQuery with its behavior.

1) Doctrine\Tests\DBAL\Logging\DebugStackTest::testLoggedQueryDisabled
Failed asserting that two arrays are equal.
— Expected
+++ Actual
@@ @@
Array
(
+ [0] => Array
+ (
+ [executionMS] => 1341871660.9651
+ )
+
)

I've submitted the PR to 2.0.x since that's where the bug is first seen and is still listed as stable on doctrine-project.org. Preferably this could also be applied to newer branches as well, or I can also submit separate PRs.

This issue is created automatically through a Github pull request on behalf of AduroIdea:

Url: https://github.com/doctrine/dbal/pull/167

Message:

driver function
To avoid ORA-03131 error, data length must be passed to oci_bind_by_name function.
This commit fixes it so now pdo_oci can be used with stored procedures and output parameters from sql/plsql.

This issue is created automatically through a Github pull request on behalf of klaussilveira:

Url: https://github.com/doctrine/dbal/pull/166

Message:

When creating a schema with tables with reserved names and relationships between them, the schema creation tool failed to write valid SQL in MySQL due to unquoted table names. For example:

CREATE TABLE `Group` (id INT AUTO_INCREMENT NOT NULL, name VARCHAR(255) NOT NULL DEFAULT NULL, PRIMARY KEY(id)) ENGINE = InnoDB;
CREATE TABLE User (id INT AUTO_INCREMENT NOT NULL, username VARCHAR(255) NOT NULL PRIMARY KEY(id)) ENGINE = InnoDB;
CREATE TABLE UserHasGroup (user_id INT NOT NULL, group_id INT NOT NULL, INDEX IDX_617A865CA76ED395 (user_id), INDEX IDX_617A865CFE54D947 (group_id), PRIMARY KEY(user_id, group_id)) ENGINE = InnoDB;
ALTER TABLE UserHasGroup ADD CONSTRAINT FK_617A865CA76ED395 FOREIGN KEY (user_id) REFERENCES User (id);
ALTER TABLE UserHasGroup ADD CONSTRAINT FK_617A865CFE54D947 FOREIGN KEY (group_id) REFERENCES Group (id);

This fix creates a small function for creating quoted foreign key table names.

This issue is created automatically through a Github pull request on behalf of Nazin:

Url: https://github.com/doctrine/dbal/pull/165

Message:

This issue is created automatically through a Github pull request on behalf of tyler-sommer:

Url: https://github.com/doctrine/dbal/pull/164

Message:

Added the necessary getters, re: doctrine/migrations#79

This issue is created automatically through a Github pull request on behalf of schmittjoh:

Url: https://github.com/doctrine/dbal/pull/156

Message:

This issue is created automatically through a Github pull request on behalf of jalliot:

Url: https://github.com/doctrine/dbal/pull/155

Message:

This issue is created automatically through a Github pull request on behalf of stof:

Url: https://github.com/doctrine/dbal/pull/151

Message:

The change in the requirements requires doctrine/common#144 to be merged and parsed by Packagist, otherwise ``2.3.*`` would not match anything.

This issue is created automatically through a Github pull request on behalf of stof:

Url: https://github.com/doctrine/dbal/pull/150

Message:

The AbstractSchemaManager expects getting the regex with its
delimiters since 8beb732fe9 but the tests were not updated.

This issue is created automatically through a Github pull request on behalf of pap-florin:

Url: https://github.com/doctrine/dbal/pull/148

Message:

This issue is created automatically through a Github pull request on behalf of Engerim:

Url: https://github.com/doctrine/dbal/pull/149

Message:

The method lastInsertId() is not defined for OCI8 Driver in Doctrine\DBAL\Driver\OCI8\OCI8Connection.php :

public function lastInsertId($name = null)
{
    //TODO: throw exception or support sequences?
}

I propose this method to handle lastInsertId for sequence :

public function lastInsertId($name = null)
 {
        // For sequence
        if (is_string($name)) {
            // We can check eventually check the presence of the sequence in the table
            // USER_SEQUENCES
            $sql = 'SELECT ' . $name . '.CURRVAL FROM DUAL';
            // will throw an exception if this sequence does not exist
            $stmt = $this->query($sql);
            $result = $stmt->fetch(\PDO::FETCH_ASSOC);
            if ($result !== false && isset($result['CURRVAL'])) {
                return (int) $result['CURRVAL'];
            }
    }
    // OCI8 driver does not provide support of lastInsertId
    return null;
}

Thanks,

This issue is created automatically through a Github pull request on behalf of franek:

Url: https://github.com/doctrine/dbal/pull/147

Message:

Add lastInsertId() support for Sequence

This issue is created automatically through a Github pull request on behalf of kimhemsoe:

Url: https://github.com/doctrine/dbal/pull/144

Message:

This issue is created automatically through a Github pull request on behalf of schmittjoh:

Url: https://github.com/doctrine/dbal/pull/143

Message:

This issue is created automatically through a Github pull request on behalf of rivaros:

Url: https://github.com/doctrine/dbal/pull/142

Message:

added support for Postgres UUID type.

This issue is created automatically through a Github pull request on behalf of dpb587:

Url: https://github.com/doctrine/dbal/pull/136

Message:

Bug fix: yes
Feature addition: no
Backwards compatibility break: no
[](http://travis-ci.org/dpb587/dbal)

This test was breaking when running under `oci8` due to Oracle's deference to upper case column names. I realize Oracle has built-in sequences and `TableGenerator` is unnecessary, in theory, but it seemed like the class should work if possible. I see pgsql is in the same boat.

Tested with:

phpunit -c oci8.phpunit.xml.dist tests/Doctrine/Tests/DBAL/Functional/TableGeneratorTest.php

Before:

PHPUnit 3.5.13 by Sebastian Bergmann.

EE

Time: 0 seconds, Memory: 10.25Mb

There were 2 errors:

1) Doctrine\Tests\DBAL\Functional\TableGeneratorTest::testNextVal
Exception: [Doctrine\DBAL\DBALException] Error occured while generating ID with TableGenerator, aborted generation: Undefined index: sequence_value

2) Doctrine\Tests\DBAL\Functional\TableGeneratorTest::testNextValNotAffectedByOuterTransactions
Exception: [Doctrine\DBAL\DBALException] Error occured while generating ID with TableGenerator, aborted generation: Undefined index: sequence_value

After:

PHPUnit 3.5.13 by Sebastian Bergmann.

..

Time: 1 second, Memory: 10.25Mb

OK (2 tests, 5 assertions)

This issue is created automatically through a Github pull request on behalf of kimhemsoe:

Url: https://github.com/doctrine/dbal/pull/139

Message:

This issue is created automatically through a Github pull request on behalf of Hounddog:

Url: https://github.com/doctrine/dbal/pull/134

Message:

Needed to remove the Paranthesis to allow for complex Regular Expressions. e.g Exclusion of tables

This issue is created automatically through a Github pull request on behalf of kimhemsoe:

Url: https://github.com/doctrine/dbal/pull/131

Message:

This issue is created automatically through a Github pull request on behalf of kimhemsoe:

Url: https://github.com/doctrine/dbal/pull/130

Message:

This issue is created automatically through a Github pull request on behalf of gedrox:

Url: https://github.com/doctrine/dbal/pull/128

Message:

This should enable caching of results retrieved by PDO::FETCH_COLUMN fetch type.
Before this exception "Invalid fetch-style given for caching result" was raised.

Should work now.

(This replaces a previous pull request #126 I failed to make right... sorry)

This issue is created automatically through a Github pull request on behalf of jimcottrell:

Url: https://github.com/doctrine/dbal/pull/127

Message:

This commit saves converted parameter values in bindValue() so the later call to the SQL logger in execute() will not pass bound objects without type information.

See https://github.com/doctrine/dbal/pull/52

When I attempt to use setCharset("utf8") with pdo_sqlite, this fails.

I know that pdo_sqlite doesn't support SET NAMES utf8, however, this should be caught by the driver (like: throwing a NotSupportedException) or simply ignore setCharset (mainly because if you allow the user to choose their backend, developers don't need to introduce special cases for driver handling).

http://msdn.microsoft.com/en-us/library/ee376925%28v=sql.90%29.aspx

I have changed a @ManyToOne relation to another entity. My workflow was this:

• Having @ManyToOne(targetEntity="Entity\EntityA") in the source
• Creating the database schema
• Changing @ManyToOne(targetEntity="Entity\EntityA") to @ManyToOne(targetEntity="Entity\EntityB")
• Calling doctrine orm:schema-tool:update

orm:schema-tool:update doesn't notice that and doesn't re-create the FK on MySQL and PostgreSQL (haven't tested others).

When I run doctrine:schema:create, doctrine throw the following exception:
[Doctrine\DBAL\DBALException]
Unknown column type requested.

( Note: their is 2 spaces between type and requested )

The problem with this error message is how much it's unspecific. Event with -v, I still have no clue about that type is wrong, and what file is concerned.
So basically, I have to manually look at all my mapping files, one by one.

Also note that doctrine:mapping:info return all OK.

EDITED: 2nd attempt to describe this - first one was confusing.

I don't understand why the setFetchMode() method of Doctrine\DBAL\Driver\PDOStatement drops any arguments passed to it. It means that PDO::FETCH_CLASS cannot be used with calls to fetch() and no warning is given until the point when the underlying \PDOStatement complains that no class has been specified.

public function __construct($username, $password, $db, $charset = null, $sessionMode = OCI_DEFAULT)
{
if (!defined('OCI_NO_AUTO_COMMIT'))

$this->_dbh = @oci_connect($username, $password, $db, $charset, $sessionMode);
if (!$this->_dbh)

}

How can i use oci_pconnect in oic8 driver?