Sane

Have any of you had experience with it? What are your takes on this combination?

I've heard there is PySQL, is that any different than MySQL with Python? What is the performance drop when using MySQL with Python, as opposed to MySQL with PHP?

And finally, is there a way to embed your MySQL database and client in to the Python Py2exe Executable, so a user does not need to install a separate MySQL client program in order to use your Python Application?

Wooh. Many questions off my shoulders. Thanks in advance.

DaWei

I've been planning to mess with it, but haven't. If you get there first, let us know. I doubt there would be much performance drop in a pro application, as the secret to speed in enterprise-sized apps is to let the DB do the grunt work and just handle the results.

__________________
Abstraction doesn't make it impossible to write bad code; it makes it possible to write superior code.
Contributor's Corner: Grumpy on C++ Exceptions DaWei on Pointers

Game_Ender

There are also python modules that map objects to databases (Object Relational Mappers, ORMs). The two I know are SQLObject, and SQLAlchemy. They both work with multiple databases on the back end (MySQL, PostgreSQL, etc.), and they handle all the SQL querying for you. You just define your classes and off you go. I have used SQLObject in turbogears, and it was pretty cool. I believe they also have the option to create hand tuned queries if you wish to increase performance.

free-zombie

I have used the MySQLdb module. It's s straightforward SQL-centric module complying with the pyhon DB API 2. It works for me.
No idea about performance

Silvanus

I've been playing with Turbogears, which uses SQLObject. I love it- an ORM makes DB programming so much easier! As to performance, it seems fast enough; I haven't benchmarked it.

EDIT: SQLObject uses MySQLdb to talk to MySQL.

__________________
:wq

Arevos

Generally speaking, MySQLdb is the defacto MySQL module for Python. It uses the MySQL library bindings (programmed in C), so it's about as fast as is possible to get. It is also fully compliant with the Python Database API Specification, which is the standard way to interface with SQL databases from Python.

In terms of performance, Python should outperform PHP, given the same platform (i.e. if you're using mod_php under Apache, a fair test would use mod_python), as PHP is interpreted (IIRC), and Python is internally compiled into bytecode before execution, which, in general, tends to operate somewhat faster than a straight AST.

That said, I'd imagine all the PHP MySQL library is codedentirely in C or C++, whilst the Python library is partially coded in Python. I'd imagine the difference between them would be small, but it would be interesting to run some speed tests.
Presumably. I can include OpenGL, SDL, GTK and wxWidgets libraries in a Py2exe excutable. MySQL should be no different.

Silvanus

I'm not so sure. MySQL isn't just a library; it runs as a service. For portable database applications like Sane's talking about, I would check out SQLite.

__________________
:wq

Arevos

Sorry; you're right. I thought Sane just wanted to include the MySQL client, but on closer inspection of the original post he wishes to include the database also (by which he presumably means the server?).

In such a case, SQLite seems the favoured option.