databases

[indigo flare]

I'd very much prefer doing this in sql instead of python

I already have the query ```sql
INSERT INTO nicknames(guild_id, user_id, {setting})
VALUES($1, $2, $3)

ON CONFLICT (guild_id, user_id)
DO UPDATE SET {setting} = EXCLUDED.{setting}```

[gusty mulch]

hi sorry for the delayed response, just tried thissql ALTER TABLE public.data ALTER COLUMN countchannel TYPE bigint[] USING countchannel::bigint;and got this errorsql ERROR: cannot cast type bigint[] to bigint LINE 3: USING countchannel::bigint; ^ SQL state: 42846 Character: 84

[austere portal]

do you already have data in the db?

[gusty mulch]

sems i do but i did deleted it yesterday, must have forgotten to save

okay I deleted the data and am getting the same error

[grim zephyr]

guys what is big int

and whats its use

[faint blade]

When you want to store very big numbers, what database do you use?

[torn sphinx]

database or data type?

[grim zephyr]

postgre

[torn sphinx]

big interger

[grim zephyr]

well i need a little help

[grim zephyr]

that am i always supposed to use client.db to connect to a postgresql

[faint blade]

Read the documentation here, as you will see BIGINT can store bigger numbers.

• https://www.postgresql.org/docs/9.1/datatype-numeric.html

[grim zephyr]

need a little help @faint blade

that am i always supposed to use client.db to connect to a postgresql

[faint blade]

What do you mean?

[grim zephyr]

i mean am i always supposed to do

self.bot.db = await asyncpg.create_pool(dsn="removed")

or can use

db = await asyncpg.create_pool(dsn="removed")

[grim zephyr]

@faint blade this

[faint blade]

You can do whatever you like, that's just setting an attribute vs assigning a variable.

[grim zephyr]

it means i can write whatever i want

??

[stray moss]

how do I establish a secure connection between my frontend JavaScript to backend database

[grim zephyr]

and am i supposed to commit in postgresql

[faint blade]

If you create a transaction yes, but asyncpg doesn't work like the other database wrappers you may have used.

[grim zephyr]

how to commit

in postgresql

[faint blade]

That is a horrible idea, someone can find the database password and username.

You should build a backend that your fronted uses, the backend would connect to the database.

[stray moss]

well if my frontend can access the backend, the user can use dev tools to access the backend and THEN access the database right?

[faint blade]

Well yes, but then they can only do things that your backend allow.

If you have your frontend complete access to the database, then they would be able to delete it.

If you create a backend, then you can restrict what someone can do.

You should implement verification

[stray moss]

hmm

I don't think it's possible to hide anything

in frontend

so even a Authorization header won't do

[brazen charm]

in client side rendering no, you cant hide anything

anything can be reverse engineered

[faint blade]

The same way someone can selfbot in Discord

But since you don't get free access to Discord's database you can at least not delete all the data

[stray moss]

hmm I wonder how

[grim zephyr]

  File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\discord\client.py", line 343, in _run_event
    await coro(*args, **kwargs)
  File "C:\Users\USER\Desktop\ZENESIS\AFK\cogs\afk.py", line 18, in on_message
    await self.bot.db.execute("SELECT user_id FROM Afk WHERE guild_id= $1", (message.guild.id,))
AttributeError: 'Bot' object has no attribute 'db'```

[thorny field]

How can I replace the userdbCheck row with newData variable?
Code:

# Buy stock command
@bot.command()
async def buystock(ctx, arg1, arg2):
    user_id = str(ctx.message.author.id)
    guild_id = str(ctx.message.guild.id)

    cursor.execute(
        "SELECT * FROM main WHERE user=? and guild=?",
        [user_id, guild_id]
    )

    do_query = """SELECT * from main"""
    cursor.execute(do_query)
 
    userdbCheck = cursor.fetchone()

    newData = userdbCheck

    if (userdbCheck is None):
        await ctx.send("You need to be registered in this server to use this command.")
    elif (userdbCheck is not None):
        if (int(arg2) < 0):
            await ctx.send("You can't buy negative shares!")
        elif (int(arg2) == 0):
            await ctx.send("You can't buy 0 shares!")

        for name, place, price in zip(["mango", "megasoft", "elecar", "soulairlines", "givida", "planetmoney", "wallmall", "chickdonald's", "gamestart", "fuber"], range(2, 11), range(0, 9)):
            if (str(arg1.lower()) == name and float(float(arg2) * stocks_list[price][1] <= float(userdbCheck[1]))):
                if (int(newData[place] > 0)):
                    newData[place] = str(int(newData[place]) + int(arg2))
                elif (int(newData[place] == 0)):
                    newData[place] = str(arg2)

[grim zephyr]

Traceback (most recent call last):
  File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\discord\client.py", line 343, in _run_event
    await coro(*args, **kwargs)
  File "C:/Users/USER/Desktop/ZENESIS/AFK/main.py", line 19, in on_message
    await bot.db.execute("SELECT user_id FROM AFK_TAB WHERE guild_id= $1", (message.guild.id,))
  File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\asyncpg\pool.py", line 530, in execute
    return await con.execute(query, *args, timeout=timeout)
  File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\asyncpg\connection.py", line 299, in execute
    _, status, _ = await self._execute(
  File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\asyncpg\connection.py", line 1625, in _execute
    result, _ = await self.__execute(
  File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\asyncpg\connection.py", line 1650, in __execute
    return await self._do_execute(
  File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\asyncpg\connection.py", line 1677, in _do_execute
    stmt = await self._get_statement(
  File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\asyncpg\connection.py", line 375, in _get_statement
    statement = await self._protocol.prepare(
  File "asyncpg\protocol\protocol.pyx", line 168, in prepare
asyncpg.exceptions.UndefinedTableError: relation "afk_tab" does not exist
```@thorny field

can you help me a little

[thorny field]

uhhh

[grim zephyr]

it is saying afk_tab does not exist

[thorny field]

is that a table or a column?

[grim zephyr]

a atable

table

in portgres

[thorny field]

i havent used that....

[grim zephyr]

@burnt turret
https://paste.pythondiscord.com/kihixufufe.py
here it is

@thorny field

[thorny field]

ok? did it give u these error before? i think ik what's the prob

[grim zephyr]

this is the first time i am using portgres

[thorny field]

ok, make sure the database is in the same folder as the program file

@grim zephyr is it?

[grim zephyr]

well can u check the code and the db for me

cause from my opinion everything is correct

@thorny field@thorny field@thorny field

[thorny field]

i haven't used prostgre but the code looks fine

[grim zephyr]

sed

@faint blade sry for the ping but can u plz help me

[thorny field]

@grim zephyr srry, idk work with dbs that much. srry i can't provide u help 😔

[grim zephyr]

Connected to pydev debugger (build 211.7628.24)
connected to database
I AM IN
Ignoring exception in on_message
Traceback (most recent call last):
  File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\discord\client.py", line 343, in _run_event
    await coro(*args, **kwargs)
  File "C:/Users/USER/Desktop/ZENESIS/AFK/main.py", line 19, in on_message
    await bot.db.execute("SELECT user_id FROM AFK_TAB WHERE guild_id= $1", (message.guild.id,))
  File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\asyncpg\pool.py", line 530, in execute
    return await con.execute(query, *args, timeout=timeout)
  File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\asyncpg\connection.py", line 299, in execute
    _, status, _ = await self._execute(
  File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\asyncpg\connection.py", line 1625, in _execute
    result, _ = await self.__execute(
  File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\asyncpg\connection.py", line 1650, in __execute
    return await self._do_execute(
  File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\asyncpg\connection.py", line 1677, in _do_execute
    stmt = await self._get_statement(
  File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\asyncpg\connection.py", line 375, in _get_statement
    statement = await self._protocol.prepare(
  File "asyncpg\protocol\protocol.pyx", line 168, in prepare
asyncpg.exceptions.UndefinedTableError: relation "afk_tab" does not exist

[thorny field]

wait wait

[grim zephyr]

ok waiting

[thorny field]

change the db name to a lowercase name

bcz it gives the error in lower case

[grim zephyr]

i tried to do that too

[thorny field]

but ur db name is upercaase

[thorny field]

oh damn

[grim zephyr]

i fixed it

@thorny field@thorny field@thorny field

fixed

!!!!

[thorny field]

how?

GOOD JOB!!!

@grim zephyr

@grim zephyr how did u fix it???

[grim zephyr]

just changed the name of the table and changed some more thing in the code

[thorny field]

ohh, cool!

[grim zephyr]

new error ```Ignoring exception in on_message
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\discord\client.py", line 343, in _run_event
await coro(*args, **kwargs)
File "C:/Users/USER/Desktop/ZENESIS/AFK/main.py", line 19, in on_message
await bot.db.execute("SELECT user_id FROM afknote WHERE guild_id= $1", (message.guild.id,))
File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\asyncpg\pool.py", line 530, in execute
return await con.execute(query, *args, timeout=timeout)
File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\asyncpg\connection.py", line 299, in execute
_, status, _ = await self._execute(
File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\asyncpg\connection.py", line 1625, in _execute
result, _ = await self.__execute(
File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\asyncpg\connection.py", line 1650, in __execute
return await self._do_execute(
File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\asyncpg\connection.py", line 1697, in _do_execute
result = await executor(stmt, None)
File "asyncpg\protocol\protocol.pyx", line 183, in bind_execute
File "asyncpg\protocol\prepared_stmt.pyx", line 171, in asyncpg.protocol.protocol.PreparedStatementState._encode_bind_msg
asyncpg.exceptions.DataError: invalid input for query argument $1: (853197674899046431,) (expected str, got tuple)

[thorny field]

lol

[thorny field]

hmm u need to convert the user_id from a tuple to str

[grim zephyr]

@thorny field (value out of int32 range)

what is this

[faint blade]

You didn't use BIGINT

[thorny field]

int variable has some bytes reserved i am guessing u are using bytes over the limit

[faint blade]

I linked you the documentation, a user ID is bigger than 32 bits

[grim zephyr]

ok

[thorny field]

convert into float?

idk

[grim zephyr]

@faint blade how to fetch data from postgre

@thorny field

[thorny field]

hmmm

let me see

[grim zephyr]

how to fetch data from postgre

[faint blade]

Look in the documentation, there should be examples of fetch(), fetchrow() and fetchval().

[harsh pulsar]

the asyncpg docs have plenty of examples

(although they are short on explanations)

(and they never actually documented the paramstyle)

[grim zephyr]

well what is the use of fetch and fetchrow

[faint blade]

fetch data from postgre

As the name suggests 😅

[grim zephyr]

i mean what is the difference

[faint blade]

The first returns a list of Row objects, the second returns a Row objects, and the last returns the first value in a Row object.

This is useful when you only select a single value. Instead of having to first get the first row from the list, then find the value in that row.

You just get it directly

[grim zephyr]

TypeError: fetchrow() missing 1 required positional argument: 'query'```

what is missing

[burnt turret]

the query

You aren't telling the database what to fetch

[grim zephyr]

is it wrong?

await bot.db("SELECT * FROM afknote WHERE guild_id = $1 AND user_id = $2", (ctx.guild.id, ctx.author.id))
    data = await bot.db.fetchrow()```

[burnt turret]

Yeah

You should pass the query to fetchrow

[grim zephyr]

well can u give a example of query

[burnt turret]

the "SELECT *..." is your query from the code you sent (that isn't the only type of query)

[grim zephyr]

ok

[burnt turret]

It basically says "give me all the rows from the afknote table"

[grim zephyr]

ok

[wind smelt]

asyncpg does it differently. Instead of calling the connection with the query then using fetchone and what not

you use fetch, fetchrow right away with your query

so it should be
data = await bot.db.fetchrow("SELECT * ...")

[faint blade]

does it better*

😅

[wind smelt]

couldn't agree more

[harsh pulsar]

the docs do explain this much 🙂

https://magicstack.github.io/asyncpg/current/api/index.html#asyncpg.connection.Connection.fetch
https://magicstack.github.io/asyncpg/current/api/index.html#asyncpg.connection.Connection.fetchrow
https://magicstack.github.io/asyncpg/current/api/index.html#asyncpg.connection.Connection.fetchval

[grim zephyr]

well i have got a error

    user = user_id[0]
IndexError: list index out of range```

[thorny field]

How can I update a row using a list in SQLite database?
Code:

# Buy stock command
@bot.command()
async def buystock(ctx, arg1, arg2):
    user_id = str(ctx.message.author.id)
    guild_id = str(ctx.message.guild.id)

    cursor.execute(
        "SELECT * FROM main WHERE user=? and guild=?",
        [user_id, guild_id]
    )

    do_query = """SELECT * from main"""
    cursor.execute(do_query)
 
    userdbCheck = cursor.fetchone()

    newData = userdbCheck

    if (userdbCheck is None):
        await ctx.send("You need to be registered in this server to use this command.")
    elif (userdbCheck is not None):
        if (int(arg2) < 0):
            await ctx.send("You can't buy negative shares!")
        elif (int(arg2) == 0):
            await ctx.send("You can't buy 0 shares!")

        for name, place, price in zip(["mango", "megasoft", "elecar", "soulairlines", "givida", "planetmoney", "wallmall", "chickdonald's", "gamestart", "fuber"], range(2, 11), range(0, 9)):
            if (str(arg1.lower()) == name and float(float(arg2) * stocks_list[price][1] <= float(userdbCheck[1]))):
                if (int(newData[place]) > 0):
                    newData[place] = str(int(newData[place]) + int(arg2))
                elif (int(newData[place] == 0)):
                    newData[place] = str(arg2)

I want to replace userdbCheck query data with newData
How can I do that?

[storm mauve]

you can use the UPDATE table SET field = newvalue WHERE field = value
(you can use anything in the WHERE clause, not only the same field you are setting btw)

https://www.tutorialspoint.com/sqlite/sqlite_update_query.htm

when it comes to executing the query, you can do it the same way you are executing the select ... where user = ? and guild = ?

[thorny field]

yea but i am using a list since the column name is a variable

[storm mauve]

what exactly do you want to update, and using which variables?

the UPDATE statement is usually used to modify already existing data in one or more rows, if you have to update the entire table then your data structure might not making much use of the advantages of the database

[grim vault]

Just a note:

    do_query = """SELECT * from main"""
    cursor.execute(do_query)```
With this part of your code you are completely undoing the where statement from the first cursor. So the next line `userdbCheck = cursor.fetchone()` will just fetch any row from the main table because there is no condition. As soon as one user is registered any user will undo the `if (userdbCheck is None):`.

[thorny field]

ohh damn, how to fix it?

[grim vault]

Just remove the two lines I quoted.

[grim zephyr]

someone help

my code is not responding at all

[torpid sky]

Is there a data type for dictionaries with POSTGRESQL

[thorny field]

will it still work like the way it used to?

[grim vault]

If you remove the lines it will really check if the context user and guild is in the main table. Before just any user/guild entry in the db would have been sufficient.

[sharp sail]

can someone explain why this may not be working?

it has area as an option

here is the data

[delicate field]

Hey @sharp sail!

It looks like you tried to attach file type(s) that we do not allow (.xlsx). We currently allow the following file types: .gif, .jpg, .jpeg, .mov, .mp4, .mpg, .png, .mp3, .wav, .ogg, .webm, .webp, .flac, .m4a.

Feel free to ask in #community-meta if you think this is a mistake.

[sharp sail]

oh

ill ss

[thorny field]

ok removed it ty

[thorny field]

cool pfp my guy

[sharp sail]

thanks

[thorny field]

np

[thorny field]

how is it color changing?

[sharp sail]

what color

[thorny field]

see it changed colors

@sharp sail ^^^

[sharp sail]

oh wow

im not sure thats cool

[thorny field]

yea pretty dope

[severe cipher]

hello! Could somebody tell me some of the best free online file storage options?

[prisma girder]

Dropbox? Google Drive? Onedrive?

[torn sphinx]

whats best database for discord.py bots?

[harsh pulsar]

sqlite for most "simple" use cases

[torn sphinx]

alr imma learn it

[hasty halo]

Hey I am geting an error Exception in Tkinter callback
Traceback (most recent call last):
File "C:\Users\shrey\AppData\Local\Programs\Python\Python39\lib\tkinter_init_.py", line 1892, in call
return self.func(*args)
File "C:\Users\shrey\Documents\RHA\main.py", line 36, in Post_Request
locals()commands
File "C:\Users\shrey\Documents\RHA\main.py", line 29, in exec2
from APP.models import Files_There
File "C:\Users\shrey\Documents\RHA\APP\models.py", line 2, in <module>
class Files_There(models.Model):
File "C:\Users\shrey\AppData\Local\Programs\Python\Python39\lib\site-packages\django\db\models\base.py", line 108, in new
app_config = apps.get_containing_app_config(module)
File "C:\Users\shrey\AppData\Local\Programs\Python\Python39\lib\site-packages\django\apps\registry.py", line 253, in get_containing_app_config
self.check_apps_ready()
File "C:\Users\shrey\AppData\Local\Programs\Python\Python39\lib\site-packages\django\apps\registry.py", line 135, in check_apps_ready
settings.INSTALLED_APPS
File "C:\Users\shrey\AppData\Local\Programs\Python\Python39\lib\site-packages\django\conf_init_.py", line 82, in getattr
self.setup(name)
File "C:\Users\shrey\AppData\Local\Programs\Python\Python39\lib\site-packages\django\conf_init.py", line 63, in _setup
raise ImproperlyConfigured(
django.core.exceptions.ImproperlyConfigured: Requested setting INSTALLED_APPS, but settings are not configured. You must either define the environment variable DJANGO_SETTINGS_MODULE or call settings.configure() before accessing settings. Its related to data bases tkinter and django. Can anyone help me

[storm mauve]

I don't think that you understand what a Database is?
Django is used to make web apps / web sites
Tkinter is used to make graphical user interfaces
Databases are used to store data (for example: sqlite, postgres, mongodb)
still, your issue is that you're not setting your settings before accessing them

[short peak]

Json best database change my mind

Sike

[storm mauve]

Well yeah, NoSQL are just Json with extra steps /s

[ripe rapids]

@remote plinth i have been learning aiosqlite

i got my blacklist to work

while testing i added myself in the db

how would i remove data from db

[remote plinth]

DELETE FROM table WHERE condition

it deletes the whole row where the condition is true

[ripe rapids]

thks cutie

[remote plinth]

[ripe rapids]

and to delete a certain thing would condition be that?

like a certain id?

[nocturne pewter]

How to connect the sqlite database file to the bot using replit web ;-;?

[remote plinth]

like WHERE column_1 = 0 AND column_2 = "test"

[ripe rapids]

thks again cutie

[remote plinth]

[nocturne pewter]

Help me plz ;-;

[remote plinth]

replit doesn't make a change, just sqlite3.connect(my_database_file)

[nocturne pewter]

K

@remote plinth how do i check my database file if it connet?

[remote plinth]

im not sure if it has a is_connected attribute

lemme read docs

!d sqlite3.connect

!d sqlite3.Connection

hmm no

lemme search docs

no again

[remote plinth]

i didn't find anything can help

[nocturne pewter]

:/

[remote plinth]

maybe try checking if db: ...

then db.close() or anything

[burnt parrot]

!e

import sqlite3

sqlite3.connect("test.db")

[delicate field]

@burnt parrot :x: Your eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "<string>", line 3, in <module>
003 | sqlite3.OperationalError: unable to open database file

[remote plinth]

test.db is not exists

[burnt parrot]

yeah ik

[austere portal]

you cant create files in sneak box

[burnt parrot]

ik

[austere portal]

but you can do sqlite3.connect(":memory:")

[burnt parrot]

k

[nocturne pewter]

Do i need to upload my database file to replit .-.?

[austere portal]

Create a database file in replit

[nocturne pewter]

K

[brave bridge]

@nocturne pewter Just make sure not to store any private data in there if the repl is public

beacuse everyone will see the file, of course

[remote plinth]

i never knew this

[austere portal]

Well, now you know it

[remote plinth]

does it reset when you restart the script?

the ":memory:"

[austere portal]

yes

[nocturne pewter]

So... how to save data in database file??

[austere portal]

Connect to the database file, do something, and commit. committing will save it

[nocturne pewter]

._.

I mean i want when i mute someone by command, bot will save their id and if they leave and rejoin server, the bot will check their id in the list if thier id in the list they still got muted role until i unmute them

[austere portal]

insert the members id to the database and commit, when a member joins (in on_member_join) select all of the ids from the database and check if the members id is in it

[grim zephyr]

@austere portal

can i have a little help

[austere portal]

Sorry, I don't know mongo db, ask your question here and some one else will answer it

[grim zephyr]

ok

[burnt parrot]

idk i use sqlite3 / aiosqlite for my discord.py databases

[grim zephyr]

what should i choose to use this as my db for my bot

[burnt parrot]

but

do

nvm idk

[torn sphinx]

i have question. i have been learning mysql and all people say that postgress more better than mysql. So should i learn postgress or continue using mysql?

thanks

[harsh pulsar]

i think postgresql is easier to use and has more useful features, for basic projects. at "big" scale there are specific reasons to choose mysql instead of postgresql, but i think postgresql is a sensible default choice

[grim zephyr]

@harsh pulsar can you help me plz and sorry for the ping

[storm mauve]

you almost definitely should look up a tutorial and follow along, setting up databases properly can be specially tricky

[austere portal]

Setting up stuff for the first time is not fun

[grim zephyr]

@austere portal

Traceback (most recent call last):
  File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\discord\client.py", line 343, in _run_event
    await coro(*args, **kwargs)
  File "C:\Users\USER\Desktop\ZENESIS\AFK\cogs\afk.py", line 34, in on_message
    result = data3[0]
IndexError: list index out of range```

its not giving me the proper value

[austere portal]

looks like data3 is empty

[grim zephyr]

            if message.mentions:
                print("executed mentions part")
                member = [member.id for member in message.mentions]
                data2 = await self.bot.db.fetch("SELECT user_id FROM afknote WHERE guild_id = $1 AND user_id = $2",
                                     message.guild.id, str(member))

                data3 = await self.bot.db.fetch("SELECT note FROM afknote WHERE guild_id = $1 AND user_id = $2",
                                     message.guild.id, str(member))
                result = data3[0]
                print("user_id = "+str(data2))
                print("note = "+str(result))
                print("member mentioned = "+str(member))  ```

here

[austere portal]

member is a list

[grim zephyr]

yes member is a list

[austere portal]

why are you casting it to str?

if you want to save a list postgres has a list data type

[grim zephyr]

maybe but why is it not getting the data3

[pearl vapor]

it is but your parameters arent correct so it cant find the data

supposedly

[austere portal]

yeah

[pearl vapor]

ofcourse we dont know the structure of your database so

and you can select several collumns in one query btw

[austere portal]

* for all columns

[grim zephyr]

here

[austere portal]

why is the user_id text?

id's are integers

[grim zephyr]

due to the brackets

though text is fine too

[austere portal]

index the data

[grim zephyr]

well they are necessary

[austere portal]

ok then

[grim zephyr]

Traceback (most recent call last):
  File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\discord\client.py", line 343, in _run_event
    await coro(*args, **kwargs)
  File "C:\Users\USER\Desktop\ZENESIS\AFK\cogs\afk.py", line 34, in on_message
    result = data3[0]
IndexError: list index out of range```@austere portal

[torn sphinx]

either is fine to learn at first, they're extremely similar for most basic tasks
i prefer postgres slightly, these are good learning sites: https://www.mysqltutorial.org/ https://www.postgresqltutorial.com/

[austere portal]

just select everyting you need in a single query

[grim zephyr]

that is giving me the whole row data

but i want only the note

[austere portal]

then data["note"]

asyncpg.Record objects are subscriptable

[grim zephyr]

wait a sec what will it do

[austere portal]

get the value of the note column

[grim zephyr]

can u plz send a example

i am getting confused

@austere portalIgnoring exception in on_message Traceback (most recent call last): File "C:\Users\USER\AppData\Local\Programs\Python\Python39\lib\site-packages\discord\client.py", line 343, in _run_event await coro(*args, **kwargs) File "C:\Users\USER\Desktop\ZENESIS\AFK\cogs\afk.py", line 34, in on_message result = data3["note"] TypeError: list indices must be integers or slices, not str

didn't work

[austere portal]

data3 is a empty list

[grim zephyr]

ok let me try once

[austere portal]

your query is unable to find anything in the db

[grim zephyr]

so how can i

??

[austere portal]

ig the ids dont exist in the table

[grim zephyr]

wth how?

they are there

[grim zephyr]

plz help bud how can i get the data

i mean nothing is working

[austere portal]

it doesnt exist is the table

[grim zephyr]

it is you saw in the ss

[austere portal]

run the sql query (with the placeholders replaced with actual values) in pgadmin and see

[grim zephyr]

wdym?

[austere portal]

run the sql query in pgadmin and see

[grim zephyr]

but how can i run in pgadmin

[austere portal]

right click "Tables" and click query tool

[grim zephyr]

ok

[grim zephyr]

its blank

nothing is there

[austere portal]

then the guild_id and member_id doesnt exist in the table

[grim zephyr]

but they are shown in the rows and data

[austere portal]

show the query you ran

[grim zephyr]

this??\

@austere portal

@austere portal

??

[austere portal]

if the query doesnt return anything that means the data doesnt exist in the table

[grim zephyr]

it is there in the rows but not showing here

but how?

[austere portal]

hmm

[grim zephyr]

!paste

[delicate field]

Pasting large amounts of code

If your code is too long to fit in a codeblock in discord, you can paste your code here:
https://paste.pydis.com/

After pasting your code, save it by clicking the floppy disk icon in the top right, or by typing ctrl + S. After doing that, the URL should change. Copy the URL and post it here so others can see it.

[austere portal]

ig the member_id passed in is not in the db

[grim zephyr]

but how

how can i save it in the db

[austere portal]

insert it

[grim zephyr]

https://paste.pythondiscord.com/ficesaxofi.py

@austere portal

can u plz check once

[austere portal]

the user id you are passing in the query doesnt exist in the db

[grim zephyr]

so how can i do that

[austere portal]

what?, if it doesnt exist it will return nothing

insert the relevant data

[grim zephyr]

ok

[grim zephyr]

@austere portal`

    await coro(*args, **kwargs)
  File "C:\Users\USER\Desktop\ZENESIS\AFK\cogs\afk.py", line 52, in on_message
    elif f"[{message.author.id}]" == user_id:
UnboundLocalError: local variable 'user_id' referenced before assignment```

[austere portal]

are you defining user_id after the elif statement?

and I cant get aiopg to work, great

[austere portal]

ok i give up

[grim zephyr]

man my file is still not working properly

[grim zephyr]

someone help

[burnt parrot]

with what

[grim zephyr]

how can i remove the record object in postgresql

[burnt parrot]

idk sorry

[storm mauve]

what is the "record" object?
a row in a table?

[grim zephyr]

well postgresql

instead of sending "note = no reason provided" its sending "<Record note='No reason provided'>"

[storm mauve]

A PostgreSQL server can have multiple databases, each of which contains multiple tables, each of which is organised in Rows and Columns

I advise taking a look at what your data currently looks like using a database explorer tool such as pgAdmin

[austere portal]

Yes, unlike normal db drivers returning tuples in asyncpg .fetch* returns a record object

[grim zephyr]

i looked into it

and everything was fine

but still the record object but why

[storm mauve]

not sure, I haven't worked with asyncpg but their documentation probably explains that in detail

[grim zephyr]

@whole widget sry for the ping but i need a littile help

[native haven]

Hello, I am reading about mongodb, and pymongo.. I was woundering, is pymongo the package I need to use to work with mongodb in async way. like is pymongo non blocking?

[brave bridge]

If you don't need to do await when you do I/O operations, it's blocking.

You can use motor or wrap the calls to pymongo into threads (using loop.to_thread)

[grim zephyr]

can u plz help me

[brave bridge]

What do you need help with?

[grim zephyr]

portgresql

[brave bridge]

What is your question?

[native haven]

oh, ok. thanks. still playing with it sync way. will take a look at that motor when i get to turn it into async

[grim zephyr]

how can i remove the record object from my value

[brave bridge]

If you want to delete something from a table, you need to execute a DELETE statement.
https://www.postgresql.org/docs/10/sql-delete.html

Is that what you want?

[storm mauve]

it sounded kinda like he wants to extract the value from the record instead of outputting the record object itself

[storm mauve]

^

[brave bridge]

oh

@grim zephyr Do you want to get the value of note from the record?

[grim zephyr]

yep

[brave bridge]

you need to do record["note"]

[grim zephyr]

ok but where and how

[brave bridge]

it's indexable by key, like a dict

[grim zephyr]

can u give a example

[brave bridge]

can you show your code?

[native haven]

sorry. remind me again.. the to_thread() is function from asyncio right?

[brave bridge]

yeah sorry, I thought it was a loop method
https://docs.python.org/3/library/asyncio-task.html#asyncio.to_thread

[grim zephyr]

!paste

[delicate field]

Pasting large amounts of code

If your code is too long to fit in a codeblock in discord, you can paste your code here:
https://paste.pydis.com/

After pasting your code, save it by clicking the floppy disk icon in the top right, or by typing ctrl + S. After doing that, the URL should change. Copy the URL and post it here so others can see it.

[grim zephyr]

https://paste.pythondiscord.com/ibumisumax.py

[brave bridge]

can you point to where you're getting a record in your code?

ah, I see

[grim zephyr]

line 34 - 36

is where i am fetching the note

[brave bridge]

asyncpg returns you Record objects with the rows that you asked for. For example, if you do SELECT foo, bar FROM baz, you'll get back a record like <Record foo=... bar=...>. You can index into it like a dictionary: record["foo"] and record["bar"].

In your case, you can do result["note"]

[grim zephyr]

what should i replace it with

[brave bridge]

Replace what?

await self.bot.db.fetch("SELECT note FROM afknote WHERE guild_id = $1 AND user_id = $2", message.guild.id, str(member))
``` this will return you a list of `Record` objects. You can access each of those objects like a dict.

Have you worked with dictionaries?

[grim zephyr]

tbh not very much

[grim zephyr]

can u tell me how to do that

[storm mauve]

you really should learn the basics instead of just asking for people to do everything for you without understand what they are doing

[storm mauve]

[grim zephyr]

sry

[grim zephyr]

i really dont know where to use it

[brave bridge]

I recommend reading this article: https://realpython.com/python-dicts/
Dictionaries are pretty important in Python, you should definitely learn how to use them.

[grim zephyr]

ok

[brave bridge]

It's fine to use more advanced libraries, but be prepared to learn a lot of stuff while implementing something

[grim zephyr]

ok

[grim zephyr]

well i am using a db so how will i use the method u said as i am not using a dict

[brave bridge]

You can index a Record by key just like a dict.

>>> d = {"note": "this is a note"}
>>> d["note"]
'this is a note'
>>> record
<Record note="this is a note">
>>> record["note"]
'this is a note'

[grim zephyr]

what are>>>

??

[brave bridge]

Have you ever used the Python console?

[grim zephyr]

no

[brave bridge]

This thing

[grim zephyr]

ye

i have seen this in vsc

but what is the use of it

[grim zephyr]

well why did u write it like this

??

[wind smelt]

to mark it as input

> indicates where your input will be

[grim zephyr]

ok

man i need to learn dict properly cause i am still not getting it

[severe cipher]

so i read that the best database to store files from dicord.py is sql? but how do i host it online?

[burnt parrot]

wdym

[severe cipher]

like suppose i want to store a file shared by a member of a server, how would i do tht?

[burnt parrot]

oh

i suggest you use sqlite3

because

its uses files

[severe cipher]

but arent sql databases stored locally?

[burnt parrot]

yes

through files

[severe cipher]

what if somebody else wants to retrieve the file?

[burnt parrot]

it saves with the bot

wait wdym

in a bad or good way?

[severe cipher]

like if those files are stored locally how would the others access that?

[burnt parrot]

you can put them a code / file sharing app

[severe cipher]

like i want ppl to store and retrieve info at any time possible while the bot is running....

[burnt parrot]

well

[severe cipher]

so like host the db

[burnt parrot]

oh

module

works with it

so like

[severe cipher]

i dont quite follow, sorry😅

[burnt parrot]

ok

sorry

[severe cipher]

np

[burnt parrot]

ok so sqlite3 is a python module which you can import and access the db files through code

so like if you wanted a custom per guild prefix

you could use that

[severe cipher]

could you elaborate this, i am a newbie in databases, sorry.😅

[burnt parrot]

ok

[severe cipher]

sorry

[burnt parrot]

ok think of the database as a box

it can store a bunch of stuff

such as fields

[severe cipher]

ok

[burnt parrot]

which can be used to organize the data

[torn sphinx]

But if you are using SQL and hosting your bot online in a ephemeral file system, keep in mind it most likely won't work how you want it to..

Databases are software that are made to store large amounts of data.

[burnt parrot]

true

[severe cipher]

umm so id be better off using a google drive api or something similar ?

[torn sphinx]

Are you hosting your bot though?

[burnt parrot]

i normaly just use docker to deploy my bots with docker so i can use my own file system

[severe cipher]

not at the moment

[severe cipher]

damn i have a lot to learn 🥲

[burnt parrot]

docker just makes deploying apps to the cloud

[torn sphinx]

If you are, then I wouldn't suggest getting a file hosting system. Instead a cloud hosting system.

[severe cipher]

ok, but are there any free ones tho?

[burnt parrot]

dockers free

docker

[severe cipher]

alright, docker it is

guys thanks a lot! 👍

[burnt parrot]

tell me if you need help with deploying it

[severe cipher]

sure thing! I cant thank you enough!

[burnt parrot]

ok

https://www.youtube.com/watch?v=z58g7_dHeMA&t=473s this guy explains it better than i would

[snow niche]

Sqlite: can I have table names with - example: CREATE TABLE IF NOT EXISTS table-name

[harsh pulsar]

technically you can, but it's not recommended. use _ instead

you would need to quote your table name in every query

[snow niche]

Ok

[grim zephyr]

!paste

[delicate field]

Pasting large amounts of code

If your code is too long to fit in a codeblock in discord, you can paste your code here:
https://paste.pydis.com/

After pasting your code, save it by clicking the floppy disk icon in the top right, or by typing ctrl + S. After doing that, the URL should change. Copy the URL and post it here so others can see it.

[untold dust]

    def load(self):
        database = self.db['database']
        cursorinfo = database.info.find()
        refinecursor = list(cursorinfo)
        self.info = dumps(refinecursor)

    def save(self):
        self.load()
        database = self.db['database']
        info = database.info
        info.delete_many({})
        for player_info in self.info:
            info.insert_one(player_info)```
question: is this the most efficient way to load and save info to mongodb?

[torn sphinx]

Hi

well

[untold dust]

hi

[torn sphinx]

I dont rlly use save and load

[untold dust]

so how do u do it?

[torn sphinx]

well

@untold dust

let me send one

player = Player(...)
level_state = Level(...)

# saving
with open('savefile.dat', 'wb') as f:
    pickle.dump([player, level_state], f, protocol=2)

# loading
with open('savefile.dat', 'rb') as f:
    player, level_state = pickle.load(f)```

#python-discussion

tell me if u need anything here.

[grim vault]

Your first step in save() is self.load() which sets self.info. Then you delete all and reinsert what you just loaded?

[grim zephyr]

any pymongo user here

[severe cipher]

hello, i am still struggling to find a way to store discord files through a bot to the cloud/online, is there any other way besides docker ?

[torn sphinx]

Huh? What’s docker got to do with storing files online?

[severe cipher]

umm tbh i dont really know 😅 😅

[torn sphinx]

So where you want to store the files? To which service?

And how do you intent to use these files again

[severe cipher]

so like i have a discord bot, and when a user posts something that they want the bot to remember/store i want it to probably store it somewhere on the cloud, so i am looking for the best option is all

[severe cipher]

its more like organising notes and automatically downloading the one which you want when you ask for it.....

[torn sphinx]

User posts something? Is they always post a file?

[severe cipher]

yes a file, an image a note something of that sort..

[torn sphinx]

If it’s file the easiest would be to just store on the local filesystem.

Or if you have a specific service in mind then it would help, because each service would have maybe a different process to upload to it

[severe cipher]

but the issue with that would be like if your system is down then the database is inaccessible right?

[torn sphinx]

Like upload to S3 is different to upload to onedrive

[torn sphinx]

Depends what kind of database. Downtime is normal and something you have to deal with. You can do things to prevent downtime

[untold dust]

me :p

[untold dust]

use repl.it

[severe cipher]

i will check it out, thanks

[severe cipher]

how do you deal with that ?

[untold dust]

you could create multiple databases

[severe cipher]

database hosting could solve this issue right/

[untold dust]

yes but self-hosting generally is worse than others

[severe cipher]

ohh

[untold dust]

bc more downtime

[severe cipher]

but where should i do this?

[untold dust]

idk you could use like mongodb and some other databases

and you could do a check like if 1 database is down use the backup database

[torn sphinx]

Everything can go down like I said. Even Amazon can go down. Just make sure to have backups and keep system up to date.

[severe cipher]

alright but i think i wasnt clear with my question, and sorry to keep asking the same thing over and over but these databases, they wouldnt be available to the bot if the systemit was created in is down right?

[torn sphinx]

This is more complex. Not something you should really worry about too much. Unless your bot is really big or downtime has major consequences on your money etc

[untold dust]

might not be that complex tho

its a pretty viable option for most ppl

[torn sphinx]

It requires replication and keeping everything in sync. If you understand how to do that then go ahead.

[severe cipher]

what i am confused about is when you create such a database it is created on your system yes? and so if the system isnt online that would mean the database isnt availabe which means the other members of the server wouldnt be able to use that feature since the db isnt available?

[untold dust]

not necessarily on your system in some cases like mongodb its hosted in the cloud

and if you host your program on the cloud using something like heroku you could get around that

[severe cipher]

ohh alrighty i did not know that!

[untold dust]

in theory

[torn sphinx]

Same issue with your bot. If discord is down or your hosting provider is down then you can’t use the bot. Stop worrying about this and just write your code.

[untold dust]

honestly it only matters if you're bot is in like a couple dozen servers or more

[severe cipher]

but isnt that rare in comparison to me shutting down my system every now and then? 😅 😅

[untold dust]

in comparison very rare

[torn sphinx]

Why would you shut down system when your trying to avoid downtime?

[severe cipher]

but what if somebody else is running this bot, they wouldnt have access to the db either right?

[severe cipher]

i dont usually run my computerlaptop throughout the day...

[torn sphinx]

Then your bot wouldn’t be running to so not sure what all this stuff your asking is about

[severe cipher]

sorry about not being clear about all this actually i understand what i need to do now, im dumb lol

And thanks a lot for the help

[austere portal]

Use a service like digital ocean/heroku to host your application

[severe cipher]

using raspberry pi aint too bad right?

[austere portal]

you can reduce the downtime by a lot that way

[severe cipher]

alright thanks

[austere portal]

That is not a bad option, but you need to run it 24/7

[severe cipher]

that shouldnt be a problem 😄 , thank you!

[fringe geyser]

e

[civic trail]

Does anyone know if I can have multiple values for a row in a sqlite3 dh

Db*

[harsh pulsar]

what do you mean by "multiple values for a row"?

[civic trail]

So we have a table called users

And then rows and columns

The rows can be

Username

Password

Hobbies

And a value of hobbies can be "basketball" which makes the hobbies row have one value

And I want to know if it is possible for the hobbies row to have 2 to 3 values

[ebon skiff]

You mean the columns?

[civic trail]

Yea

[ebon skiff]

You can use an array inside the database or just seperate them with a comma

Both are pretty bad

[brave bridge]

@civic trail
Separating values with a comma isn't a good idea. It will make it really hard and inefficient to filter/group by hobbies in such a system. What you want to do is create a separate table, called user_hobbies, like so:

CREATE TABLE user_hobbies (
    user_id INTEGER REFERENCES users(id),
    hobby TEXT NOT NULL
);

So you'll have a situation like this:

# users
id | name  | age
---+-------+---
 1 | alice | 16
 2 | bob   | 25

# user_hobbies
user_id | hobby
--------+------
      1 | programming
      1 | tennis
      1 | drawing
      2 | drawing
      2 | dancing
      2 | beer
      2 | videogames

this is called a many-to-many relationship, where you have many users, each of whom has many (0 or more) hobbies

[ebon skiff]

It is that is the sad part is

but then def maria

[brave bridge]

If you set up a table like that, you'll be able to create an index: sql CREATE INDEX idx_user_hobbies ON user_hobbies(hobby); An index is a separate structure, similar to a Python dict, that allows you to quickly look up rows: py idx_user_hobbies = { "programming": [1] "drawing": [1, 2] } For example, if you make a query like sql SELECT DISTINCT u.id, u.name, u.age FROM users u JOIN user_hobbies h ON h.user_id = u.id WHERE h.hobby = 'programming'; to find all users who have programming as a hobby, and SQLite won't scan through the entire table -- it will look up which user IDs have a programming tag in the index, and it will only look up those rows.

[torn sphinx]

Bad how? They just follow the python specification for database drivers. Just like other libs

Do you have any particular issue with it? Most drivers follow the same interface so usage will be the same

Use the official connector

[civic trail]

@brave bridge thats awesome

But in my blog post website commas will be ok

Since im just adding the users who upvoted

So can you tell me how to separate with commas?

[civic trail]

Btw im talking about this

[torn sphinx]

                            filecontent = filecontent.replace(i, str(ctx.author.id) + "|" + str(nowAmount) + "|" + str(int(timedingdong) + 60 + "|" + int(bank) + "|"))

its for my discord bot but it comes up with hundreds of errors

i added bank and it broke

it saves in a file

but it deletes instead of replaces

[wraith tundra]

I'm not sure about this, but is there a way to make a database with OOP? Sorry if this is the wrong channel.

[civic trail]

With oop does he mean object orientated programming

[wraith tundra]

yep

[wraith tundra]

yep

[shell ocean]

look into SQLAlchemy

Peewee

and Django's ORM

[pliant spire]

    class User(Base):
        __tablename__ = "users"

        user = Column(type_=String, primary_key=True)


    Base = declarative_base()
    engine = create_engine(f'sqlite:///{get_project_root()}/db/main_db.db')
    session = sessionmaker(bind=engine)
    session = session()
    # @todo, does this mean all my models have to live here?
    Base.metadata.create_all(engine)
    session.add(User(user='hello_world_l33t'))
    session.commit()
    session.close()

Hi guys, I was wondering how Base.metadata does its job?

I'm wondering if i'll need to keep all my models in the same file as the Base.metadata.create_all(engine) call

Is there a way to import that, so the Base knows what all my tables look like?

this is sqlalchemy + sqlite

[gentle apex]

is SQL considered a database?

[harsh pulsar]

no SQL is a language that many databases use for querying

[pliant spire]

anyone please?

[storm mauve]

not sure, but can't you just access the sqlite_schema table?

[low shoal]

so I have this error with asyncpg

PostgresSyntaxError: syntax error at or near "ON"

while using this query

INSERT INTO afk (user_id, last_seen, reason)
    VALUES ($1, $2, $3) RETURNING *
ON CONFLICT (user_id) DO UPDATE
    SET last_seen = EXCLUDED.last_seen,
        reason = EXCLUDED.reason
    RETURNING *;

Some research showed that the problem may be because of old postgresql version but I'm using 13.2
or is that query wrong?

well removing the returning * worked

although that raised another error

but why can't I use returning there

ok so now I am getting prolog InvalidColumnReferenceError: there is no unique or exclusion constraint matching the ON CONFLICT specification

so my on conflict statement is probably wrong

[harsh pulsar]

you had RETURNING * in two different places, that was probably the original error

[low shoal]

o

[harsh pulsar]

the new error means that there's nothing to be in "conflict" with

what's the definition for this table?

it looks like user_id is supposed to be the primary key, but for some reason isn't

[low shoal]

CREATE TABLE afk (
   last_seen TIMESTAMP,
   user_id BIGINT NOT NULL,
   reason TEXT,
   PRIMARY KEY (user_id)
);

well it should be the primary key

[harsh pulsar]

huh

[low shoal]

yeah probably the on conflict is wrong because wherever I see an example of on conflict it is done differently

[low shoal]

I fixed it by changing PRIMARY KEY to UNIQUE

now it works as it should

[harsh pulsar]

that shouldn't be necessary. weird

double check the docs

[wraith tundra]

thanks

[zealous parcel]

Hi, i use aiomysql but there is a delay in it How do I know the reason for this comment?

[brave bridge]

Can you show your code?

[zealous parcel]

http://prntscr.com/1mvpprs

http://prntscr.com/1mvq297

[brave bridge]

@zealous parcel What do you mean by 'delay'?

[zealous parcel]

When calling it delay until it responds only when I use this aiomysql

[brave bridge]

Have you tried testing it without the bot?

[zealous parcel]

no

[brave bridge]

Then you should do that and measure the time

Are you using a local database or a demote database? If it's a remote database, there could be a pretty big latency

I had hundreds of milliseconds of latency with ElephantSQL

[zealous parcel]

When I turn on the bot it works normally but this happens after a while of running

[zealous parcel]

loaclhost

[zealous parcel]

@brave bridge

[brave bridge]

can you perhaps look at the connection list in your database?

https://stackoverflow.com/questions/1620662/mysql-see-all-open-connections-to-a-given-database

[next sun]

hello

I am facing the following problem

psycopg2.OperationalError: server closed the connection unexpectedly
    This probably means the server terminated abnormally
    before or while processing the request.

on trying to fetch the data I need.

it happens that I've tried a few things out of what I found googling but nothing seems to work

though, I have found this github discussion

https://github.com/psycopg/psycopg2/issues/829

[brave bridge]

@next sun Maybe you have too many open connections?

[next sun]

and there is a dude that has this comment:

I'm having the same issue. For now as a workaround had to create wrapper around connection object with used methods to hook them. In cursor() method had to do a SELECT 1 "ping" check and reconnect if needed to return a valid working cursor.

but I can't understand how I could implement this workaround

can anyone help me?

that would probably fix the problem temporarily at least

[next sun]

humm

when running htop

I see not script

basically I setup a cronjob to run a script everyday

[brave bridge]

I don't think htop will show that. Can you run SELECT * FROM pg_stat_activity; on the database?

[next sun]

you are right that the script in the ends hangs kind of waiting for something (only when it fails on the db side)

[next sun]

I think I can

let me see...

[next sun]

I can run

but....

what is important in the table?

[brave bridge]

Run this ```sql
SELECT
pid
,datname
,usename
,application_name
,client_hostname
,client_port
,backend_start
,query_start
,query
,state
FROM pg_stat_activity
WHERE state = 'active';

that should show the active connections

or remove the WHERE clause and see all the connections

[next sun]

is only shows one active

the interesting thing is that when I run the script manually second time, no error happens

without the where clause it shows about five or six

@brave bridge is it ok if I kill connections

[zealous parcel]

http://prntscr.com/1mw7rtx

[jade shell]

(ping) is there a good tutorial how how to make a data base for a discord bot

[prisma girder]

Have you seen pinned messages?

[gentle roost]

Is there a convention for naming the return of a database search?

I've been doing

vals = cursor.execute("SELECT * from ...", (...,)).fetchall()

vals was temporary but I started using it everywhere out of habit

[prisma girder]

results?

Depending what you receive

It can be users, products, accounts and so on

[gentle roost]

Gotcha.

[mortal light]

I was looking to add a database to a project and came across psychopg2 driver. But psycopg3 also exists.

Should I use psychopg3 instead of psycopg2?

[austere portal]

Wait what psycopg3 is there

[mortal light]

I was just going through the basics of psycopg2. How to connect and disconnect to the database. Haven't really done anything substantial, so I can use psycopg3 if it's better.

[mortal light]

Yes, https://github.com/psycopg/psycopg

it's written in python

[austere portal]

Thanks I'll check it out, I've been using psycopg2 for all this time

[mortal light]

Do you use psycopg2 on regular basis?

[austere portal]

Yes

I have been using that for all my projects which use the postgresql database

[mortal light]

At a company or on personal projects?
Just trying to gauge if more people are still using psycopg2 at companies, then I could start off with psycopg2.
And once the project is done, I dont mind refactoring it to psychopg3.

[brazen charm]

generally psycopg2 is still the more popular setup

just because psycopg3 is still a WIP

[austere portal]

it has async support, nice

[brazen charm]

asyncpg still king though

psycopg2 has such a legacy of weird feature choices and bad performance in places asyncpg will likely never be moved off being king as an async driver

[mortal light]

Yeah, I think you have the right assessment, mostly because it takes companies few years to adopt a new technology.

[brazen charm]

depends on the company, and generally modules like that are pretty simple to interchange in a good setup

[austere portal]

Yeah, but I like the pep, it is nice

[mortal light]

Are queries written the same way in PostgreSQL and in psycopg2?

[brazen charm]

sorta

psycopg runs with it's own placeholder setup

sql syntax the same otherwise

postgres runs with the $n syntax, psycopg2 does mysql style %s placeholders

with a awful mix of encouraging string formatting

[brave bridge]

yeah, psycopg has to do its SQL injections query formatting, because $N isn't in the DB API 2 spec

[mortal light]

I dont know either, so I guess I should be using the keyword pscopg2 when I look up how something should be done, instead of postgreSQL.

[brazen charm]

yeah

well

depends, sql logic -> postgres, placeholder behavour -> psycopg2

why my execute many being slow? -> psycopg2

[harsh pulsar]

"db-api" and "pep 249" are also good search keywords

!pep 249

[delicate field]

**PEP 249 - Python Database API Specification v2.0**

Link

Status

Final

Created

29-Mar-2001

Type

Informational

[harsh pulsar]

also i think psycopg3 is out, or at least in some kind of beta state https://www.psycopg.org/psycopg3/

[brave bridge]

why my execute many being slow? -> pip uninstall psycopg2 && pip install asyncpg

[mortal light]

so is "pep 249"/ "db-api" a wrapper for most common databases? So queries would be consistent across different databases

[harsh pulsar]

not a wrapper, but a pattern that database libraries are generally expected to follow

this pep is why most python database libraries have very similar APIs

[brave bridge]

but are still incompatible because they have different query syntaxes

[harsh pulsar]

except asyncpg, which deliberately deviates from the standard https://magicstack.github.io/asyncpg/current/faq.html#does-asyncpg-support-db-api

[mortal light]

So if someone wrote a database library for python, then they would follow the conventions mentioned in db-api, so developers would find it easy to use this library?

Oh ok I see.

[harsh pulsar]

precisely

[dense barn]

im trying to change my default value of one column from 0 to 1 but it does not update it, ive faced this issue twice now, anyone know why?

[austere portal]

How are you updating the value?

The query?

[dense barn]

im just changing it through pg admin

[austere portal]

An alter query will do the job

[dense barn]

i did that to

didnt work

ALTER TABLE ONLY pokesetspc ALTER COLUMN card_amount SET DEFAULT 1
``` is what i did

[austere portal]

Hmm

Any error?

[dense barn]

nope

said its successful

[austere portal]

How are you inserting the data?

[dense barn]

just updating it

UPDATE pokesetspc...

[austere portal]

The default value of a column will only apply if you are inserting a new row without the column specified

[dense barn]

bruh

so does that mean only the new columns will have 1 ?

[austere portal]

Yeah

[dense barn]

bruv

how do i change the old ones then?

[austere portal]

Update all of the rows then

[dense barn]

oh well, ig its gonna be 0 then

[forest crystal]

https://cdn.discordapp.com/emojis/768543607492706314.png?v=1&size=40

[mortal light]

When using psycopg2, we would only do conn.commit() if it's some sort of writes/deletes? For read we dont have to do conn.commit()?

[brave bridge]

run conn.commit() when you need to commit a transaction.

I think the context manager automatically does it, but I could be wrong, you'll have to check

[mortal light]

When you say "context manager automatically does it", do you mean psycopg2 driver automatically does commit?

I was dropping a table and it didnt drop it without commit

There's an option to enable auto commit if that's what you are referring to.

[brave bridge]

Are you using psycopg's context managers?

[mortal light]

Sorry, I dont know what that is. Can you post a code snippet and I might get an idea what you mean.

[brave bridge]

A context manager is something you use in a with statement

[mortal light]

oh no, I am not

[brave bridge]

You should look into this: https://www.psycopg.org/docs/usage.html#with-statement

Are you familiar with transactions in SQL?

[mortal light]

Taking the concept of regular context manager of with to read and write and extending that to psycopg2, I remember reading it's better to use with in case reads and writes are too big, and with has closing file descriptor baked in.

I am looking at the with psycopg2 you linked. No, I'm not familiar with transactions in SQL. I just started using databases.

A transaction is a logical unit of work that contains one or more SQL statements So SQL transactions would execute multiple sql statements/queries.

[brave bridge]

Imagine that you want to save this object to the database:

@dataclass
class Post:
    id: str
    title: str
    content: str
    tags: list[str]

This is one way to do it: ```py
def save_post(connection, post):
connection.execute(
"INSERT INTO posts (id, title, content) VALUES (%s, %s, %s)",
(post.id, post.title, post.content)
)
connection.executemany(
"INSERT INTO post_tags (id, tag) VALUES (%s, %s)",
[(post.id, tag) for tag in post.tags]
)

app = SlowAPI()

@app.post("/posts")
def insert_post(request):
post = post_from_json(request.json())
if post is None:
return 422
conn = create_db_connection()
try:
save_post(conn, post)
finally:
conn.close()
return 201, {"url": f"/posts/{post.id}"}

It's no big deal here, but that could violate some other important business rules.

So what you want to do is either execute both statements or none of them. That's called an "atomic operation" -- it either executes fully or doesn't execute at all.

[mortal light]

"INSERT INTO posts (id, title, content) VALUES (%s, %s, %s)", do we always use "%s" regardless of what data type is being stored in the database?

[brave bridge]

Yeah, I think so

You could also do %(foo)s, %(bar)s and then pass a dictionary like {"foo": 1, "bar": "baz"}

I haven't used psycopg, only asyncpg, so I'm just looking at the docs right now as to how to use transactions

[mortal light]

ok I'll set up basic methods to create, drop table and insert data and then run it through more data and see what needs to get adjusted.

[brave bridge]

that's why you want to use transactions:

@app.post("/posts")
def insert_post(request):
    post = post_from_json(request.json())
    if post is None:
        return 422
    conn = create_db_connection()
    try:
        with conn:
            with conn.cursor() as cur:
                save_post(cur, post)
    finally:
        conn.close()
    return 201, {"url": f"/posts/{post.id}"}

in psycopg a transaction is automatically created when you execute the first statement.

(or so it seems from the docs)

[mortal light]

After you mentioned context manager and I looked at the example, I didnt think I could use it because how I am structuring my code by creating different methods, but I think I should still be able to use context manager.

[brave bridge]

...this is equivalent to

    conn = create_db_connection()
    try:
        with conn:
            with conn.cursor() as cur:
                save_post(cur, post)
                conn.commit()
    finally:
        conn.close()
        conn.rollback()

[mortal light]

Yeah I was looking at something similar!!

[brave bridge]

what are you writing, if that's not under an NDA?

[mortal light]

Making an API call and then storing some data to the database, so whenever someone runs the program for the next 24 hours it would fetch data from the database instead. And after 24 hours drops the table and makes the call to api again.

It would be 5000 rows with 4 columns each.

[brave bridge]

the reason with doesn't close the connection is probably that it has to play nicely with pools

@mortal light Why not use SQLite?

it can take 5000 rows just fine

[mortal light]

I thought about that, but the program would run from different computers.

[brave bridge]

Do you actually need to mutate the database from different servers?

[mortal light]

What do you mean by "mutate the database"?

[brave bridge]

why do you want several computers to interact with the database directly?

[mortal light]

If I used SQLite, then the data fetched from api call would be local to that computer. If another user runs the program from a different computer then it would make call to api again.

[brave bridge]

Are you exposing the database directly to the users of your application?

[mortal light]

And I am trying to rely less on this third party API.
Database and the program is for internal use.

[brave bridge]

That sounds like a bad idea, what if someone sends DROP TABLE things;?

I would wrap it in an web server (flask/fastapi) and then users would make requests to that API

Sounds like less than 100 lines of code, and that would prevent a lot of bad things

[mortal light]

How would someone run DROP TABLE things?

It's a local python program/script with database.

So only people who have this program would be able to run it.

[brave bridge]

You said that different computers would have access to the database, right?

maybe I'm misunderstanding the context

[mortal light]

Yeah so the program makes request to API, response data gets stored locally(as of now in excel *facepalm).

If I run the program/script it will make api call and store data on this computer for 24 hours.
If some other person at company runs the program/script from their computer, it will also make call to the api and store data on their computer for 24 hours.

So anytime someone new runs the program, it will make call to the api. Third party API is slow and has call limit.

So creating a small database seemed like a reasonable way to address this issue. This way one computer would make api request and then every other computer after it will be fetching data from the database when they run the program.

[brave bridge]

@mortal light What I'm suggesting is instead of exposing the database directly to all the other computers, expose an API, so someone can do

requests.get(f"{YOUR_SERVICE}/bananas?id=42")
``` and get the data they need

[mortal light]

Oh I see what you mean, even though the program is being used internally, it's still a better approach to create an api for it?

[brave bridge]

Yeah

because Alice opens a phising link and opens up a hole into your system, then Bob spills coffee on the keyboard which accidentally sends DROP TABLE things; etc.

and it's easier to interface with

[mortal light]

And other computers would also be able to make post request if the data is older than 24 hours.

So then I would have to create a server and host this api?

[brazen charm]

you would need to do that regardless

but yes

[mortal light]

Yeah I have an RDS instance running for this postgres database.

ok I will think about this.

[brave bridge]

@mortal light You could just update the data in the background, right?

e.g. with a systemd timer https://wiki.archlinux.org/title/Systemd/Timers

and while it's updating you either refuse to serve responses, or just serve outdated data if that's fine

[proven arrow]

Just curious but how long have you been trying to solve this problem? I’ve seen you ask about how to go about this few months ago and your still figuring it out?

[brave bridge]

btw shouldn't excel handle 5000 rows just fine?

just attach an API to excel 😄

and I'm only half joking

[mortal light]

Yeah I asked something about database and sqlite couple of months back, but I had to focus on rest of the project, so now I am getting to database.

[mortal light]

lol I asked something a while back which mentioned excel.
And someone said, "Of course, some people use excel in their projects. It's poor mans database" lol and from their I found out about SQLite to store data locally.

But if I get the chance to go back to those projects, I guess I would rewrite the excel portions to utilize SQLite, so there isnt a concern of someone opening one of the excel files.

[torn sphinx]

i have been download postgresql and this happening. Please help me!

[dense barn]

how would i match a column from database 1 to database 2? And then if the matching is correct i want to take something from a certain column of databse 1 and insert it to database 2. any idea how i would do that?

and by matching i mean match a lot of things

i can try searching on the internet, lmao never though of that

[dense barn]

ok, well that was unexpected, i didnt think it would be this easy but i got it.

[green iris]

When migrating data in Django, is there a way to ser the get natural key behavior? I am unable to import my user data as some names are blank but I want to serialize on the user's email address instead

[narrow moth]

I need to offer the ability to search (read-only) a postgres database via discord but it feels like I'm getting way off track.

Full SQL is too complex for people to learn (realistically) so I ended up making a tokeniser that might be able to build queries it feels like it's just an extra five steps in a weird direction before getting back to just updating a dictionary and using if statements to parse the input into a pyscopg.sql function.

Do I keep going down this road or is there a better way?

[tribal cargo]

Hello, I try to initialize an flask_sql Database. But I get some: can't set attribute error. I'm not sure whats the problem. From what I read it should work. I tried it in an extra task to figure out the problem:

models:

tracestack

Is anybody firm with this issue?

[snow niche]

anyone know sqlite?

[torn sphinx]

What is your question

[snow niche]

can i connect to db like this?

class Logs(commands.Cog):
    """📔 Set custom mod-logs channel."""
    def __init__(self, bot):
        self.bot = bot
        self.conn = aiosqlite.connect("db's/moderation.db")
        self.c = self.conn.cursor()
    @_set.command()
    @commands.has_permissions(manage_channels=True)
    @commands.bot_has_permissions(manage_channels=True)
    async def modlogs(self, ctx, channel:discord.TextChannel):
        """Set a mod-log channel using this command."""
        conn = await self.conn
        cursor = self.c
        await cursor.execute("INSERT OR REPLACE INTO mod_logs(channel_id, guild_id) VALUES (?, ?)", [channel.id, ctx.guild.id])
        await conn.commit()
        
        await ctx.send(f'Mod-log channel set to {channel.mention}.')

[torn sphinx]

TIAS

Try it and see

[snow niche]

my internet is not that good, so thats why it took so long to reply

[torn sphinx]

Should work

I think you need to await the connection, though

[snow niche]

i did await self.conn

[torn sphinx]

Hmm, then try it and see

[snow niche]

ok just want to make sure

[snow niche]

is the INSERT OR REPLACE correct? i want it so like when some one uses this command once again and changes the mod logs channel... i want it to delete the original mod-log channel and replace with the new one, does this work that way?

[austere portal]

Await the connect

You cant use await out side async functions so use loop.run_until_complete

[torn sphinx]

it may work, not sure about this function

[torn sphinx]

yeah

[torn sphinx]

await aiosqlite.connect("db's/moderation.db")

[pure sleet]

why are you running create_all on before request?

you do realize that that is going to run everytime a request is made basically

[brave bridge]

I'm using PostgreSQL, and I want to store blog posts that have a list of media objects as their content, for example: ```js
[
{tag: "header", text: "Using JSON as a database in 5 easy steps"},
{tag: "sub_header", text: "Denial"},
{tag: "paragraph", text: "..."},
{
tag: "horizontal_split",
left: {tag: "sub_header", text: "Anger"},
right: {tag: "sub_header", text: "Bargaining"}
},
...
]

JSONB also supports indexing

[tribal cargo]

O thought it's to create the database

[pure sleet]

you only need to create it once

[tribal cargo]

Yes, that's why it's in the before first request method. When launching it will run once as long as the server is running

[pure sleet]

before_request runs before each request, not the first

[tribal cargo]

But it is before_first_request(). Let me check that

[pure sleet]

ok i see my bad. but still

just run it while the app is not in production to create the database

sqlalchemy will still connect to it

even if it exists

[tribal cargo]

Still doesn't work without the code

Even if i run from Terminal

[jaunty galleon]

Is it possible to get the datatypes of each column in asyncpg using some query?

Like that it would return (guild_id bigint, something text)

(column_name datatype)

[narrow moth]

not sure about asyncpg, but for psycopg:

    col_details = []
    try:
        cur = conn.cursor()
        cur.execute(sql.SQL("SELECT column_name, udt_name, character_maximum_length, column_default FROM INFORMATION_SCHEMA.COLUMNS WHERE table_name LIKE {}").format(sql.Literal(table_str)))
        col_details = cur.fetchall()
        cur.close()
    except psycopg2.Error as e:
        print(e)
    return col_details```

(oh, but also: sql.Literal() is one of the things I'm meant to fix... )

[wind swan]

I have a question about database design

I will get straight to my use case.

I am making a discord game, about db. So oc there are transformations in it

but I need to somehow store the data of the transformations, including the limiting rules (discord rule that you have to own on order to transfrom)

so that thing that came to my mind is:

the problem here, that if I have couple of rr (required rules) for the same transformation

lets say, to get ss3, you have to own ss1 and saiyan

And I can't store multiple values in the same call since NF1

So I need help with that

[pure sleet]

when does this error occur? when you run create_all?

[narrow moth]

what database are you using?

[wind swan]

MySQL

Someone told my to not follow NF1, because it's not must

but im not sure about that

[narrow moth]

hmmm, I am not sure I can help with this very much (haven't used MySQL) but it sounds like you'd want an array type.

Postgres does nice arrays. I'm looking pretty hard at them lately and they look awesome.

[wind swan]

yes, oc. but the problem is NF1. I can't have 2 objects on same cell

[wind swan]

So ☝️

[narrow moth]

Personally, I'm in a place where I'd promote getting a working concept up and running before worrying about the finer details. Not in the case of injection attacks and such, but maybe work out how to comply with NF1 later.

let me mull a minute.

[wind swan]

um...

[narrow moth]

I'm not sure. Optimisation is beyond my power level.
Even if you had table with boolean columns for each transformation, it just sounds like a lot more back and forth.
If transformations are linear (i.e. SS2 requires only SS1, SS1 requires only Saiyan, etc) then it's one value per cell, but you'd be traversing a fair bit.

I'm sure there's a way to do it, it sounds like it's something people would want to do pretty frequently.

[wind swan]

thanks a lot for trying to help man!

I'll think about that and ask other people

[snow niche]

ok

also btw should i await the cursor??

[austere portal]

Yes

[narrow moth]

Good luck mate.

[austere portal]

Normally it's used in a async with statement

[snow niche]

alright

[snow niche]

so do u think this will work?

[austere portal]

Don't think insert or replace is valid

[grim vault]

SQLite does have INSERT OR REPLACE INTO ...

[austere portal]

Oh didn't know that, thanks 😃

[pure sleet]

normalize it to like 3rd or fourth form

[wind swan]

yes, but how

all the NF requiring not having more then one object in a cell

[pure sleet]

you split the table to confine with the normal forms

you can use relationships to keep the data linked

[wind swan]

ill think on somth

[quasi aurora]

I have used a destructor to close a database in OOPs
Do i also have to delete all the other variables declared in the class or will python take care of that?

[brave bridge]

You don't need to delete any objects yourself (in fact, you can't do that), Python will delete them if nobody can reach them, That's called "garbage collection".

You shouldn't normally use __del__, it's more of a last resort or a hack. It's not defined in the language when the __del__ method is called, or whether it's called at all. Instead, you should close the database connections explicitly, like you close files.
Depending on the library, it will look somewhat like this:

def transfer_balance(sender_id: int, receiver_id: int, amount: int):
    if amount < 0:
        raise Exception("Negative balance")
    with connect_to_db() as conn:
        with conn.transaction():
            if conn.fetchval("SELECT balance < ? FROM users WHERE id = ?", (amount, sender_id)):
                raise Exception("Not enough money")
            execute("UPDATE users SET balance = balance - ? WHERE id = ?", (amount, sender_id))
            execute("UPDATE users SET balance = balance + ? WHERE id = ?", (amount, receiver_id))
            conn.commit()

[quasi aurora]

I mean, what if I have to close a database after a signal to end the execution of code is sent

[brave bridge]

Connection objects and other resources already define __del__ so that they will be closed in case of a termination.

[quasi aurora]

oh

[brave bridge]

Additionally, the OS will free up all the resources you've taken

[quasi aurora]

ty

[quiet birch]

I have a query on mongodb

I have an ID and type key

I want them to be unique together but not on their own

so I can have multiple of Type and multiple of ID even though they are the same

but if ID = 1727417 and type = 0 I should not be able to re insert it due to duplication

[harsh pulsar]

@quiet birch https://stackoverflow.com/a/14693445 ?

[quiet birch]

uh

[sharp kindle]

anyone got a good resource for designing a datbase schema

[proven arrow]

Any good guide on normal forms will help you achieve the basic principles. But there is no universal designing guide. What may seem good for one project might not be good for another. Best is to understand your problem your working on and then work around that.

[dim sluice]

pymongo.errors.ServerSelectionTimeoutError: Does anyone know this error?

^ I had changed ip for some reason and that is the reason

[hasty hinge]

SELECT *, ROW_NUMBER() OVER(ORDER BY  xp DESC) as rank
FROM
  usuarios
WHERE
  guild = "a"
  AND id = "a";

What is wrong on the syntax of this query? I'm using MySQL 8

You have an error in your SQL syntax; it seems the error is around: 'rank FROM usuarios WHERE guild = "a" AND id = "a"' at line 6

It was because of rank

Looks like you can't use vars named rank in MySQL 8 (it used to work on MySQL 5)

[grim vault]

RANK (R); added in 8.0.2 (reserved)

It became a keyword.

[whole pendant]

i want to user to change username but when i want to write the username that already existed the clean_username method not working and thus giving IntegrityError at /users/profile-edit/ UNIQUE constraint failed: users_user.username
HOW TO QUERY IN clean_username SO THAT IS SHOW ONLY USERNAME ALREADY EXIST ValidationError

class UserEditFrom(forms.ModelForm):
    username = forms.CharField(label="",max_length=254,widget=forms.TextInput(attrs={'id':"user-name",'type':"text" ,'class':"form__input"  ,))

    def clean_username(self,*args, **kwargs):
        username = self.cleaned_data.get("username")
        if User.objects.filter(username=username).exclude(username=username).exists():
            raise forms.ValidationError('username is alraedy taken try another one')
        return username

[austere portal]

There is a built in UserEditForm class in django, use that it will automatically do the validation

[whole pendant]

I HAVE MAKE COSTOM USER MODEL

[austere portal]

ok

did you do py class Meta: model = User fields = (...)?

[whole pendant]

THE BUG IS IN clean_username method

[austere portal]

You are filtering and excluding

[whole pendant]

yes this query i want

[austere portal]

so .exists() will return false

[whole pendant]

hmm

[keen sundial]

Fyi mongo bug recently fixed:

You are receiving this email because one or more of your Atlas clusters was previously running a MongoDB version that was susceptible to a recently identified bug.

The bug, which impacts MongoDB versions 4.4.7, 5.0.0 and 5.0.1, may allow some insertions to succeed even though they should fail with a duplicate key error against a unique index. This bug does not allow documents to be inserted with duplicate _id values.

[ancient folio]

Mongo DB is the best for py?

[brazen charm]

thats a very broad question

generally the language has nothing to due with the suitability of a database

[tardy pike]

Can you elaborate?

[brazen charm]

the suitability of a database generally depends on
a) the structure of the data, e.g. is it relational, unstructured, etc...
b) the scale of the application e.g. sqlite is great for small projects but cant scale
c) often cost, cloud services charge different amounts for different databases and offer different management packages

[tardy pike]

b) What is the limit of sqlite in terms of size of a project?

[brazen charm]

depends on the application, but you're limited to being single threaded with heavy locks so performance degrades significantly when you get larger tables or need concurrency

[snow niche]

But then can't we create more files?

[tardy pike]

Great input, thanks. Any suggestions for DB that is a good choice in terms of scalability? Financial data.

[snow niche]

For each topic a file

[remote plinth]

@ripe rapids (sorry for ping, had to)

[brazen charm]

generally Postgres and / or Scylla (Cassandra) is a pretty good choice depending on scale and amount of writes needed

[tardy pike]

I will have a look into those. It would be a good amount of writes needed .

[supple quail]

Hello guys, how to use SQLite in memory, any good article with example to teach me how, thx

[brazen charm]

literally the same as normal sqlite except you connect to :memory: rather than a file path

[supple quail]

Oke noted 👍

[fierce dirge]

How to update json file with python ?

[pure sleet]

what do you mean?

[fierce dirge]

I want to update json file, opened with open() ?

I have some data in json and I want to add something in it

[pure sleet]

convert it to a dictionary and then update that, once done, just dump the dictionary to a json file

[fierce dirge]

But when I update, updated data is not reflected in the file

[fierce dirge]

Means open file, get dict, close file, update dict, then open file in write mode and write the dict and close file ??

[pure sleet]

i guess, use the json library in python

[fierce dirge]

Yes I used it

[pure sleet]

don't just write the dict, convert it to json first, but i guess you know that

[fierce dirge]

Yes

But I want in same file

[pure sleet]

bro, you write to the file you want to change no? so it's overwritten with the updated data

[fierce dirge]

Oh yea

Yes*

I get it

Thanks buddy

[pure sleet]

sure

[ripe rapids]

had to what.

[lone island]

asyncpg returning json data as string, how can i avoid that and return dict instead ??

[brazen charm]

probably use a custom Record factory / Record class

[lone island]

more details ?

[brazen charm]

actually scrap that

see https://magicstack.github.io/asyncpg/current/usage.html#example-automatic-json-conversion

[lone island]

not sure thats goona help me

[brazen charm]

you wanted automatic json conversion

see the example that tells you how to do that

[lone island]

so after configuring that, i can return json as dict using fetch ?
res = await conn.fetchval('SELECT $1::json', data)
they used ::json extra

[remote plinth]

to ping

i closed all my dms...

[brazen charm]

thats just explicitly say the type to postgres

[remote plinth]

so i was searching for yours

[brazen charm]

you need it when it's without a schema

if you're column is already defined as a json datatype then you dont need to type cast

[lone island]

uh

My issue is with regular fetch and how it is translated into the json using standard python utility.

[brazen charm]

you will likely want to use $n::json though when inserting data

[brazen charm]

yes

if you set a custom encoding and decoder asyncpg will use that when inserting and fetching the data before it's returned to you

[lone island]

im still not sure how to use them

on fetch

[brazen charm]

you dont touch anything with fetch

you just set the type codec on the connection

and it should do everything else without needing to touch anything

[lone island]

so all i need to do is to set that codec and it'll return json and no str ?

[brazen charm]

it should do yes

[lone island]

alright thanks imma try it then

[ripe rapids]

why dont u just search my name

[remote plinth]

spaces.

[ripe rapids]

u cn still search it

[remote plinth]

ah k