I couldn't get it to work

But I found workaround for ordering by object id instead of attribute of object

Hello, is it possible to somehow use a pony with a singleton-class?

Depends on your singleton class.

Okey, this is a tough one. I just migrated my tables to basically group similar `telegram_id`s. So I used the following code to figure out if a channel is marked NSFW (1 = allow NSFW, 0 = spoiler, -1 = don't show NSFW) or Safe For Work: (1 = allow SFW, 0 = spoiler, -1 = don't show SFW) channels = orm.select(channel for channel in ChannelSource if not channel.disabled and channel.allow_nsfw == CHANNEL_RATING.ALLOW.value or (channel.allow_nsfw == CHANNEL_RATING.SPOILER.value and channel.allow_safe in (CHANNEL_RATING.SPOILER.value, CHANNEL_RATING.HIDE.value))) Now I try to rewrite that to figure out if a TelegramChannel that as well. Basically NSFW can be treated as max aggregation, because if one source is nsfw, you have to expect nsfw. For safe it is the other way around, min that is if all allow safe.

I think I can get the allow_nsfw part done with attribute lifting, somehow similar to that: channels = orm.select(channel for channel in TelegramChannel if not channel.disabled and max(channel.channel_sources.allow_nsfw) == CHANNEL_RATING.ALLOW.value or (max(channel.channel_sources.allow_nsfw) == CHANNEL_RATING.SPOILER.value and min(channel.channel_sources.allow_safe) in (CHANNEL_RATING.SPOILER.value, CHANNEL_RATING.HIDE.value))) But now I don't know how I can filter out everything which is disabled…

Can I use … if all(<condition> for source in channel.channel_sources) ?

In other words, iterate over the Set attribute?

If it works, I could get away with channel: TelegramChannel source: ChannelSource channels = orm.select( channel for channel in TelegramChannel if not (show_disabled or channel.disabled) # hide disabled channels and channel.channel_sources # hide channels which have no sources and all( # all sources must conform to: source.disabled # either be disabled, and thus not influence the result or source.allow_nsfw == CHANNEL_RATING.ALLOW.value # must allow nsfw or ( # NSFW is spoilered, but safe is also spoilered or not even shown. source.allow_nsfw == CHANNEL_RATING.SPOILER.value # NSFW is allowed, but spoilered and source.allow_safe in (CHANNEL_RATING.SPOILER.value, CHANNEL_RATING.HIDE.value) # and safe is also only spoilered or even hidden ) for source in channel.channel_sources ) )

Hi, I trying to make another filter in filter but I can't be successful. I have two table, Transaction and Log. My main table is Transaction and I use the Log table for operations such as creation, updating, deletion on the Transaction. Tables are; class Transaction(db.Entity): id = PrimaryKey(int, auto=True) ...other_columns... logs_set = Set(Log, cascade_delete=True) class Log(db.Entity): id = PrimaryKey(int, auto=True) time = Required(datetime, default=lambda: datetime.utcnow()) type = Required(int) ...other_columns... transaction_ref = Optional('Transaction') class TYPE: CREATE = 1 UPDATE = 2 DELETE = 3 My goal is to receive Transaction not marked as deleted. The queries I tried is cash_box.transactions_set.filter(lambda t: t.logs_set.filter(type=Log.TYPE.DELETE).count() == 0) I'm getting this error: File "/home/myilmaz/Clouds/GitHub/myilmaz-web/venv/lib/python3.7/site-packages/pony/orm/sqltranslation.py", line 1401, in wrapper return method(monad, *args, **kwargs) TypeError: filter() got an unexpected keyword argument 'type' How I can do this. Can you help me?

I believe type is reserved Python keyword

Maybe this cause your issue?

I believe type is reserved Python keyword

No, it isn’t

Ah, you right

Its function https://docs.python.org/3/library/functions.html#type

Yes, I tried filtering with another column and I got an error again

Shouldn't there be lambda function?

oh, thank you so much. That was the problem.

Can I use … if all(<condition> for source in channel.channel_sources) ?

No, currently you can't because SQL does not have FOR ALL clause. But you can express it with double nested negation with NOT EXISTS clause. The resulted query will look pretty complex and non-intuitive (it is hard to understand an exists clause with double negation), but it should work. Something like orm.select( channel for channel in TelegramChannel if not channel.disabled and channel.channel_sources and not exists( source for source in channel.channel_sources if not ( source.disabled or source.allow_nsfw == ALLOW or (source.allow_nsfw == SPOILER and source.allow_safe == SPOILER) ) ) ) ) Basically, "select a channel for which does not exist a source which does not satisfy necessary conditions".

I'm a new, help me please with select. I need all records from P with specific R and specific M with order by time. Class R id = PrimaryKey(int, auto=True) name = Required(str) prop = Required(str) ps = Set('P') ms = Set('M') Class M id = PrimaryKey(int, auto=True) mid = Required(str) rs = Set(R) Class P id = PrimaryKey(int, auto=True) time = Required(datetime) v1 = Optional(str) v2 = Required(str) r = Required(R)

I understand if you want P related to specific R, as they have relationship. But what do you mean by "with specific M"? P is not connected to M, and if you already have "specific R" I don't see how you plan to further restrict it to "specific M"

«specific mid»

PonyORM created a table m_r but I cant link to it in the request

It may be something like: specific_r = R[some_id] query = select(p for p in P if p.r == specific_r).order_by(lambda p: p.time) or specific_m = M.get(mid=some_value) query = select(p for p in P if specific_m in p.r.ms).order_by(lambda p: p.time)

hi, good: one problem: pony.orm.core.BindingError: Database object was already bound to MySQL provider, why is solution?

hi, good: one problem: pony.orm.core.BindingError: Database object was already bound to MySQL provider, why is solution?

You cannot do db.bind(...) if you already bound db object to some specific database

now: TypeError: No database converter found for type <class 'pony.orm.core.PrimaryKey'> help mw, please!

why is solution?

You probably specified something like attr = Required(PrimaryKey), don't do that

It may be something like: specific_r = R[some_id] query = select(p for p in P if p.r == specific_r).order_by(lambda p: p.time) or specific_m = M.get(mid=some_value) query = select(p for p in P if specific_m in p.r.ms).order_by(lambda p: p.time)

Added condition specific_m = M.get(mid=some_value) query = select(p for p in P if specific_m in p.r.ms and p.r.name).order_by(lambda p: p.time) but in result It shows me everything in P without specific_m. In table m_r I have a several records m.id — r.id 1 - 2 2 - 3 2 - 2 2 - 4 So, in result I see all m.id where r.id = 2 (for example). I need only r.id = 2 where m.id = 1 (mid is str in table M)

You probably specified something like attr = Required(PrimaryKey), don't do that

very thanks!!

Added condition specific_m = M.get(mid=some_value) query = select(p for p in P if specific_m in p.r.ms and p.r.name).order_by(lambda p: p.time) but in result It shows me everything in P without specific_m. In table m_r I have a several records m.id — r.id 1 - 2 2 - 3 2 - 2 2 - 4 So, in result I see all m.id where r.id = 2 (for example). I need only r.id = 2 where m.id = 1 (mid is str in table M)

Your question does not look entirely clear to me. Probably you want something like: select( (p, m) for p in P for m in p.r.ms if p.r.name == foo and m.mid == bar )

Your question does not look entirely clear to me. Probably you want something like: select( (p, m) for p in P for m in p.r.ms if p.r.name == foo and m.mid == bar )

I tried this variant earlier but it’s show an error assert name_path not in sqlquery.tablerefs

I tried this variant earlier but it’s show an error assert name_path not in sqlquery.tablerefs

Can you show me some reproducible example code for this bug?

only this query in db_session or you need a model code ?

Probably a query should be enough

But it is better to have a code which can be executed

Hi guys! Is there any known issues with using the "coalesce" function within @property?

I'm having an issue when Im reading a Decimal value from an object and its fine with the correct resolution

like:

print(str(a.value)) # prints 0.00

(value is a Decimal field with 2 digits after the comma)

but if I do: print(a.property_value) # this @property uses a coalesce print(a.value) # prints 0.0000 even though the resolution is 2

actually I clarified the issue

If I do: print(a.fk_object.value) # prints a string from that object print(a.value) # prints 0.0000

it has nothing to do with @property and @coalesce it seems

but just with decimal when playing with other objects that have variables with higher decimal values

I'm not sure this is the actual reason. Can you try this sequence: print(str(a.value)) print(a.value) print(a.fk_object.value) print(a.value)

that works

if I read the value before accessing the foreign key

Yup

so what you said works, but this fails: print(a.fk_object.value) print(a.value)

even just doing print(a.value) before works

and makes it so that the second print is correct

then the value is correct

it seems like the computer gremlins are at it again, this a very odd issue haha

It looks strange. What database do you use?

Is it SQLite?

Postgres

the fk_object has a variable that is decimal and of resolution 4

could be related to that

hi guys, might be a simple question. How do we do fuzzy search in pony orm ? I have a q=tshirt and would like to search my Product db entity with 3 fields (name, description, tag). In django, it would be like q((name__icontains='tshirt')|(description__icontains='tshirt ')|(tag__icontains='tshirt'))

Probably param = tshirt select(x for x in X if ( param in x.lower() or param in x.description.lower() or param in x.tag.lower() ))

But for real user queries it is probably better to use full-text-search queries

Pony does not have out of the box support for full text search queries, but you can add necessary tables manually

thank you very much!

Just playing around with orm.migrations and it works great for me! 👍 But it took me quite a bit of searching to figure out how to install it: https://github.com/ponyorm/pony/issues/16#issuecomment-400161683

Maybe it could be useful to advertise this more?

Migrations are not ready yet. orm-migrations is previous attempt. New branch is migrations_dev. Both are incomplete at this moment...

You can try to use it on your own risk, but some important features may be missing

Yeah, I saw migrations_dev. It seems good enough for my simple application.

Anyways, just saying thanks for the work you already did on it!

You're welcome!

I'm really happy the pony is still alive and kicking :) Keep it up!

😊

Pony is so awesome. It doesn't get in my way.

hey guys 🙂 @metaprogrammer may I ask you what library you used for the connecting svg-lines/arrows? https://editor.ponyorm.com/user/pony/PhotoSharing/designer it looks fabulous 🤌

hey guys 🙂 @metaprogrammer may I ask you what library you used for the connecting svg-lines/arrows? https://editor.ponyorm.com/user/pony/PhotoSharing/designer it looks fabulous 🤌

Just SVG And JavaScript Thank you!)

so you wrote that algo on your own?

Yep

that’s wild =)