An error occurs in set_attribute when using PDO SQLSRV

[SakiTakamachi]

Description

The following code:

<?php

// my env
$pdo = new PDO('sqlsrv:server=mssql-server;database=test;', 'test_user', 'p@ssw0rd');
$pdo->setAttribute(PDO::ATTR_STRINGIFY_FETCHES, false);

Resulted in this output:

Fatal error: Uncaught PDOException: SQLSTATE[IMSSP]: An invalid attribute was designated on the PDO object. in /var/www/html/test.php:5
Stack trace:
#0 /var/www/html/test.php(5): PDO->setAttribute(17, false)
#1 {main}
  thrown in /var/www/html/test.php on line 5

A change here is triggering the error.
#11622

The essence of the problem is that PDO SQLSRV is throwing an error instead of returning false when set_attribute fails.
microsoft/msphpsql#1468
laravel/framework#47937

I've been working on this issue for about 20 days now and the hotfix for PDO SQLSRV has not yet been released and things are looking a little shady.

However, since there are many people who are in trouble, I thought about creating a PR for a workaround with php-src. What do you think?
#12039

remarks:
I've checked all the PDO drivers provided by PECL and found that only one, PDO SQLSRV, gives an error under the same conditions.

PHP Version

PHP 8.1.22+, 8.2.9+, 8.3.0+

Operating System

No response

[SakiTakamachi]

@Girgias
Since you were reviewing #11622, I would be happy if you could give me your opinion.

[Girgias]

I don't think we should be monkey patching around a database driver, even though in some sense, the previous PR was monkey patching the MySQL driver. But this has good reasons, as the MySQLnd driver supports differing behaviour.

That PDO attribute has existed since at least PHP 5.3, and was probably somewhat of a bug to not propagate this down to the underlying driver.

Moreover, there might be reasons for the underlying driver to actually warn/throw. I don't understand why we should be monkey patching when seemingly the only thing preventing this issue to be fixed is for a multi-billion corporation to release a new version of their driver without doing any work as you provided the fix.

People who are affected by this should be pressuring Microsoft in releasing a new driver, not pressuring us.

[SakiTakamachi]

@Girgias
Thank you for your feedback!

I don't understand why we should be monkey patching when seemingly the only thing preventing this issue to be fixed is for a multi-billion corporation to release a new version of their driver without doing any work as you provided the fix.

Yes, honestly I completely agree.
I was proposing a fix that I wasn't satisfied with, but I was a little relieved to find someone with the same opinion.

I haven't heard back from the PDO SQLSRV lead yet, so I'll wait for his response first.

Thank you for your precious time.

[SakiTakamachi]

I am closing this issue and the associated PR.
I hope it never opens again.