Are you looking for an #online #database #modeling #tool with #SQL generation?
With the latest updates (e.g. modeling of #association #classes), BESSER is your ideal tool to #design and #generate your databases in many different "flavours" :
and even #SQLAlchemy
Ready to give it a try? No installation required! 
https://editor.besser-pearl.org/
Did I already mention is completely 
and #opensource?
I have a #Python codebase using #sqlalchemy Core -- *NOT* ORM -- for its data layer, and I'm currently not happy with the way it's doing data for testing. Who has a pattern for this that they actually like?
I already have ephemeral per-test-function isolation of the database, so that's taken care of.
If this were ORM I'd just use something like factory-boy, but the fact that it's Core and doesn't really map neatly to objects the way ORM stuff does rules that out. I think ideally what I'd like is some sort of declarative JSON file format or whatever that lets me provide a mapping of table names to lists of rows to insert into them, and then a pytest fixture that lets me say "load that JSON file for this test". Has anyone written something like that? Do I get to go write my own?
Just added SQLAlchemy from @zzzeek to our list! From raw SQL nightmares to elegant Python ORM - you've transformed how an entire generation of developers works with databases.
Thank you for this masterpiece of engineering!
#Python #SQLAlchemy #OpenSource
SQLite vs PostgreSQL: A Performance Showdown in Real-World Applications
In a battle between SQLite and PostgreSQL, one developer's quest for optimal performance reveals surprising insights about database capabilities. As applications scale, understanding the nuances of th...
https://news.lavx.hu/article/sqlite-vs-postgresql-a-performance-showdown-in-real-world-applications
As an example, I have started using #SQLAlchemy recently in my daily work. Starting out with a new software library can take a long time reading documentation to get started using it properly.
ChatGPT has been very helpful here. I might ask it something like: "I need to represent these two different kinds of objects that have these properties, and I need to be able to track many-to-many relationships between them through foreign keys. How can I achieve this using SQLAlchemy and ORM models?"
@veronica #sqlalchemy has been nothing but trouble in our team. Lazy loading makes you have more transactions that last longer and also leads to the n+1 query problem. And I hate the query constructor to me it’s harder to understand than plain old SQL. We’re stripping it off to get back down to the #psycopg2 layer.
@veronica Give a try to #peewee - much more simple than #SQLAlchemy !
Using #SQL directly instead of #SQLAlchemy reminds me of the time I stopped using #TailwindCSS and started using "vanilla" #CSS. Not as challenging as I thought and actually quite enjoyable.
Miguel Grinberg: What's New in SQLAlchemy 2 @ PyConIe 2023
At first I just wanted to play with some parts of the code like the ranking function. However it didn't take long before I decided that I wanted to build something more and, why not, play with some things I normally don't get to play with in #python like #async stuff and the #SQLAlchemy #ORM
Thus, winzig was born:
https://github.com/dnlzrgz/winzig
