Enhancement: add support for $driverOptions

Use correct classname in exception message

Fix a Typo

Update tests/Doctrine/Tests/DBAL/Functional/Schema/PostgreSqlSchemaManag...

Use platform specific datetime format for assertions in WriteTest

DBAL-645: Oracle Platform: Cannot list table columns from a different schema

Using user_col_comments in the INNER JOIN when referencing all_tab_columns causes no results to be returned.

Skip MasterSlaveConnectionTest for vendors other than MySQL

 * test case to assert the connection timeout is set
 * test case to assert on exceptions thrown
   * wip, as it seems impossible to trigger 'false' from mysqli_options

 * push error handling down to when it occurs
 * WIP, as I am unable to trigger false currently

 * move supportedDriverOptions into setDriverOptions
 * make setDriverOptions private
 * support \MYSQLI_SERVER_PUBLIC_KEY on 5.5.0+

 * fix typo
 * remove fluent interface

 * introduced $_supportedDriverOptions to validate what is passed into the object
 * added `setDrivrOptions()` which sets the array and throws exceptions

Remove unused Table::$_idGeneratorType

Fix reserved keyword conflicts in TemporaryTableTest

Fix Index construction

Update CreateSchemaSqlCollector.php setting default namespace/schema nam...

supportsSchemas() was already added (see last commit).
So changed expectation from 1 to any.

Added the supportsSchemas method in the mock because it's being used now.

The current implementation uses the schema name as namespace name.
Some platforms don't support schema's, but a namespace name still needs to
be present. Added namespace name 'default' when there is no default schema
name.

specified stable 2.4.* releases of doctrine/common in composer.json

SQL Server Paginator bug fix

Apperently the doModifyLimitQuery function does not only cover SQL
generated by ORM (Makes sense to use DBAL as seperate component). Therefor
it can happen that a column name does not include the table and so the old
logic was valid anyway. (line 867-869)

Fixed the unit tests to use real column names instead of aliases.

1. combine the regexes into one. The only thing that needs to be captured
is the alias. Used a non-capturing group for "AS " to avoid confusion and
have cleaner regex. It tries to match "Table.Column Alias" or
"Table.Column AS Alias". (line 861)
2. No fallback on line 865 like "$column['table'] . '.' .
$column['column']" this wouldn't make sense as this can not be possible
when it's a wrapped query. Better not insert some random SQL that's gonna
confuse the hell out of everybody.
3. Removed "($column['hasTable'] ? $column['table']  . '.' : '') .
$column['column']" because the original regex always tries to match the
table name. Therefor hasTable will always be true or the entire function
fails anyway. (this references line 868 now) (this was unit tested on
mysql, postgresql and sql server and did not increase errors/failures)

default schema name is public for postgresql not default. Added debug
statements to test travis-CI.