Porting to mysqli from mysql

[code green]

I am still using the mysql extension rather than the mysqli.
Moving to mysqli would be realively simple in my case because all mysql_ procedural calls are wrapped in a class MySqlDB().
As are mssql_ procedural calls and both inherit from an abstract class that handles common functionallity baseDB().

So to switch I need to edit a few lines in the MySqlDB class.
At least I think it is that simple.
But as I already have a class inheriting from a base class what is the best way to implement the mysqli extension?

By inheriting mysqli from MySqlDB class?
Instantiating mysqli from MySqlDB class?
Or competely abandon MySqlDB and re-write hundreds of scripts?

[Atli]

Hi.

I would just replace the calls to the mysql extensions with calls to the mysqli extension. In your case it should be as simple as modifying a few lines in your MySqlDB class, right?

This is the point of encapsulation and abstraction in OOP; to be able to modify the inner workings of a class without affecting the code already using it.

Unless you want to have both versions: MySQLDB and MySQLiDB, or something.

[code green]

I have only glanced at mysqli tutorials, but don't you have to create a mysqli object to use the extension?
I have also noticed complaints of a performance loss from switching.
Can't corroborate this but idly googling for performance pulls a few articles

[Atli]

You can use mysqli in two ways, using the mysqli object itself, or the mysql-like functions. The functions were created to ease the transition from the old mysql extension, I believe.

Simply adding a "i" to your old mysql function call should work for most of them.
Just check out the manual entries. They always list a procedural function and a OOP method for each of them.

As to the performance loss, I don't know. I've never tested it myself, but the few articles I've seen about this didn't really worry me.
Anyways, the real power of the Improved MySQL extension is in the prepared statements.