I tried with sqlite and memory DB like in example - same result

Strange, I never encountered such behaviour before

What python version you using?

3.11

Did you override select by any chance?

Oh, wait. Pony doesn't support Python 3.11 yet.

😂

Course, that I study - worse that I think)

What version can advice?

3.10 or earlier ?

"Can be installed on python 2.7 or 3" - official doc, all info about versions)

3.6 - 3.10 included

Yeah should be updated

is the idea to support 3.11 in the future? Also I tried to load a table into a dataframe using the following query but that errors out @orm.db_session def load_table_into_df(table): attrs = [x.name for x in table._attrs_with_columns_] q = orm.select((getattr(d, a) for a in attrs) for d in table) # << This fails df = pd.DataFrame(q, columns=attrs) return df

is there an easy way to select all _attrs_with_columns_ in a query?

something equilvalent of select * from table

ended up with something like this @orm.db_session def load_table_into_df(table): attrs = [x.name for x in table._attrs_with_columns_] data = [el.to_dict() for el in table.select()] df = pd.DataFrame(data) return df

I can suggest using some migrations manager which runs raw sql

Are migrations still on the Pony road-map for the next future? Or is there a recommended migration manager (postgresql).

I really liked firebase

Migrations are almost done for postgres almost 2-3 years, to be honest.

How to add multiple objects to a table at once?

At this point it would be not fair to give any promises about its release. Since we gave a lot of them.

How to add multiple objects to a table at once?

You mean bulk insert?

You mean bulk insert?

yes

https://stackoverflow.com/questions/51655297/python-pony-orm-insert-multiple-values-at-once/51670740#51670740

https://stackoverflow.com/questions/51655297/python-pony-orm-insert-multiple-values-at-once/51670740#51670740

How can I get these values ​​from the list? exp: ids = [1,2,3,4,5,6,90] User.select(lambda u: u.id in ids) # this not work

Migrations are almost done for postgres almost 2-3 years, to be honest.

So it's available on a branch?

So it's available on a branch?

Yes, it's a bit outdated in terms of python support. I guess it stuck at 3.9 version or smth. If I remember correctly the branch is orm_migrations. It supports most of common actions, not tested on some big composite changes. Mostly tested on PostgreSQL and SQlite. Should also somehow work for both Oracle and MySQL.

How can I get these values ​​from the list? exp: ids = [1,2,3,4,5,6,90] User.select(lambda u: u.id in ids) # this not work

Not sure why... Seems like it should

Yes, it's a bit outdated in terms of python support. I guess it stuck at 3.9 version or smth. If I remember correctly the branch is orm_migrations. It supports most of common actions, not tested on some big composite changes. Mostly tested on PostgreSQL and SQlite. Should also somehow work for both Oracle and MySQL.

Ok I think I will try to update it to Python 3.10. Do you want a PR on it if can figure it out?

That would be nice. Actually migrations and python bytecode support don't intersect in general, so maybe simple merge\rebase should do the job.

That would be nice. Actually migrations and python bytecode support don't intersect in general, so maybe simple merge\rebase should do the job.

ok

Yes, it's a bit outdated in terms of python support. I guess it stuck at 3.9 version or smth. If I remember correctly the branch is orm_migrations. It supports most of common actions, not tested on some big composite changes. Mostly tested on PostgreSQL and SQlite. Should also somehow work for both Oracle and MySQL.

Is the migrations_dev branch also related or is this a different thing?

Oh yeah, you're right. migrations_dev is the branch you need. orm_migrations is a super old migrations approach which was declined.

Not sure why... Seems like it should

I would like to make a selection by certain id

How can I get these values ​​from the list? exp: ids = [1,2,3,4,5,6,90] User.select(lambda u: u.id in ids) # this not work

Try this: ids = [1,2,3,4,5,6,90] values : list["User"] = [ User.get(id = id) for id in ids ]

Try this: ids = [1,2,3,4,5,6,90] values : list["User"] = [ User.get(id = id) for id in ids ]

this many iterations)

this many iterations)

Yes 😅 I don't know if PonyOrm support other method if you found tell me to learn also..

Good day everyone, I just started using ponyorm my question is can I separate my models into different files

Like all the entity that belongs to use in one file and all the model that belongs to post in another file

Whats the problem uve met?

Please... I want to ask if ponyorm is still being maintain😕, I want to use it for a big project, that will be hard for me rewrite. Thanks

According issues from the pony repo the cockroachdb is no longer supported.

I want to know how true

And lastly is there a migration tools

Honestly, it is maintained badly. I had almost no resources to work on PonyORM lately. I believe it should change now, as I can work on Pony again. But if I were you, I'd take this with a grain of salt. I suggest you see if we could release a new version of PonyORM with Python 3.11 support in a few days, and if not - consider some other options to be safe. I seriously plan to return to work on Pony again. But it is better not to rely on specific dates. The support of CockroachDB should be restored in the same release as adding Python 3.11 support. Regarding migrations, I cannot promise specific dates for them.

Honestly, it is maintained badly. I had almost no resources to work on PonyORM lately. I believe it should change now, as I can work on Pony again. But if I were you, I'd take this with a grain of salt. I suggest you see if we could release a new version of PonyORM with Python 3.11 support in a few days, and if not - consider some other options to be safe. I seriously plan to return to work on Pony again. But it is better not to rely on specific dates. The support of CockroachDB should be restored in the same release as adding Python 3.11 support. Regarding migrations, I cannot promise specific dates for them.

I've used Pony as my main orm since 2020 and still having couple projects in production using it. However I've switched to async code for my projects around year ago and it wasn't suitable for them so I had to switch to other orm. None of them provides such flexibility and comfort I've had with Pony and it still hold a special place in my heart. I'm looking forward to see what you can do with and how you can improve this outstanding orm and wish you luck!

Any (coding) help would be appreciated

Honestly, it is maintained badly. I had almost no resources to work on PonyORM lately. I believe it should change now, as I can work on Pony again. But if I were you, I'd take this with a grain of salt. I suggest you see if we could release a new version of PonyORM with Python 3.11 support in a few days, and if not - consider some other options to be safe. I seriously plan to return to work on Pony again. But it is better not to rely on specific dates. The support of CockroachDB should be restored in the same release as adding Python 3.11 support. Regarding migrations, I cannot promise specific dates for them.

Thanks

But if the message is >aa

Is this an appropriate place to ask a question regarding relationships and cascade delete?

Is this an appropriate place to ask a question regarding relationships and cascade delete?

yes, absolutely

@akozlovsky Is it possible for you to take a look at this PR: https://github.com/ponyorm/pony/pull/677 It is a fix for this issue: https://github.com/ponyorm/pony/issues/675 For the record, it is an issue we are hitting in production. I have deployed a work-around in my system, but it would be nice to slove this properly.

@akozlovsky Is it possible for you to take a look at this PR: https://github.com/ponyorm/pony/pull/677 It is a fix for this issue: https://github.com/ponyorm/pony/issues/675 For the record, it is an issue we are hitting in production. I have deployed a work-around in my system, but it would be nice to slove this properly.

Thank you for attracting my attention to this very important issue and the PR. I'll try to handle it today. Need to think a bit about the proposed fix and whether it is completely correct

@akozlovsky Is it possible for you to take a look at this PR: https://github.com/ponyorm/pony/pull/677 It is a fix for this issue: https://github.com/ponyorm/pony/issues/675 For the record, it is an issue we are hitting in production. I have deployed a work-around in my system, but it would be nice to slove this properly.

The identified reason is correct, and the provided test is very helpful, but the proposed solution does not look complete; it reduces the number of cases when the bug appears but does not eliminate them. It does not fix the situation when a new function with the same id has the same code object as the garbage-collected function (but contains different closures). Also, it can lead to some memory garbage remaining uncollected. The correct solution should use weak refs and discard translators that refer to garbage-collected functions. I'll try to come up with it.

The identified reason is correct, and the provided test is very helpful, but the proposed solution does not look complete; it reduces the number of cases when the bug appears but does not eliminate them. It does not fix the situation when a new function with the same id has the same code object as the garbage-collected function (but contains different closures). Also, it can lead to some memory garbage remaining uncollected. The correct solution should use weak refs and discard translators that refer to garbage-collected functions. I'll try to come up with it.

Great, thanks for looking into this. Let me know if you need any assistance.

Great, thanks for looking into this. Let me know if you need any assistance.

Sure!

hey guys quick question - is it possible to create a constraint where say I have two columns but not both can be filled in - one or the other have to be filled in

smth like this https://stackoverflow.com/a/960591/1764089

If your database allows that, you can manually add constraint: alter table ... add constraint ... And also add before_insert/before_update handlers to the entity with a check

thank you

Hello everyone, I just switched from Python 3.10 to Python 3.11.2. Suddenly get an error in a simple select query. Does somebody has any idea ?

Hello everyone, I just switched from Python 3.10 to Python 3.11.2. Suddenly get an error in a simple select query. Does somebody has any idea ?

Python 3.11 is not yet supported, unfortunately. @metaprogrammer said he wanted to fix this as soon as he gets around to it.

Python 3.11 is not yet supported, unfortunately. @metaprogrammer said he wanted to fix this as soon as he gets around to it.

Oh thank you. I should have read that correctly :) !

how can I choose the schema I create my tables in?

seems like this is all I need to do? class Person(db.Entity): _table_ = ("my_schema", "person_table") name = Required(str)

Yes

do the a if b else c expressions work now? - I tried them and got a "this expression is too complex" message https://stackoverflow.com/questions/16115713/how-pony-orm-does-its-tricks#comment59802214_16118756

I works if you supply query as a string instead of a generator code select((x.a if x.b else x.c) for x in X) - does not work select("(x.a if x.b else x.c) for x in X") - works

Ok got it

yeah that worked - though got a weird error when I wrap it with orm.min it returns '\n' but if I wrap an orm.select() with orm.min it returns the correct value

Also getting an ERDiagramError: Entity Table already exists when I define define tables in different schemas ie model1.py from utils import db class Table(db.Entity): _table_ = 'schema1', 'table' model2.py from utils import db class Table(db.Entity): _table_ = 'schema2', 'table'

i'd expect this to work but it complains

when i later do import model1 import model2

posted the question here if easier: https://stackoverflow.com/questions/75602565/in-ponyorm-do-i-need-to-declare-classes-with-different-names-even-if-they-belong

Why you are trying to re-create same entity?

Ah, different schemas, i see

It says about class name, it should be different

the class names should be different right?

Yes