postgresdb3 2.1.1

High-Performance Sync and Async PostgreSQL ORM for Python

🐘 PostgresDB3 ORM

postgresdb3 - Python tilida yozilgan, PostgreSQL ma'lumotlar bazasi bilan ishlashga mo'ljallangan, o'ta tezkor va yengil ORM (Object-Relational Mapping) kutubxonasi. U sinxron (psycopg2 orqali) va asinxron (asyncpg orqali) ishlash rejimlarini to'liq qo'llab-quvvatlaydi hamda Django-ga o'xshash qulay sintaksisni taqdim etadi.

---

📌 Mundarija

1. O'rnatish (Installation)
2. Ma'lumotlar Bazasiga Ulanish
3. Modellarni Yaratish (Models)
4. Maydonlar (Fields)
5. Validatorlar (Validation)
6. Model Metodlari va Hooklar (Hooks)
7. CRUD Operatsiyalari
8. QuerySet API (Qidiruv va Filtrlash)
9. Aloqalar bilan ishlash (Relationships)
10. Migratsiya Tizimi (CLI va Kod orqali)

---

1. O'rnatish (Installation)

Loyihangizga kutubxonani o'rnatish uchun:

pip install postgresdb3

Eslatma: Agar siz sinxron dvigatelda ishlamoqchi bo'lsangiz, tizimingizda libpq-dev (Linux) o'rnatilgan bo'lishi kerak. Aks holda psycopg2-binary kutubxonasini alohida o'rnatib olishingiz mumkin.

---

2. Ma'lumotlar Bazasiga Ulanish

PostgresDB3 sinxron va asinxron ulanishlarni alohida dvigatellar yordamida amalga oshiradi. Ulanishlar hovuzi (Connection Pool) avtomatik tarzda boshqariladi.

Sinxron ulanish (PostgresDB)

from postgresdb3 import PostgresDB

db_sync = PostgresDB(
    database="my_db",
    user="postgres",
    password="my_password",
    host="localhost",
    port=5432,
    minconn=1,   # Minimal ulanishlar soni
    maxconn=20,  # Maksimal ulanishlar soni
    echo=True    # SQL so'rovlarini terminalga chiqarish (Debug)
)

Asinxron ulanish (AsyncPostgresDB)

from postgresdb3 import AsyncPostgresDB

db_async = AsyncPostgresDB(
    database="my_db",
    user="postgres",
    password="my_password",
    host="localhost",
    port=5432,
    min_size=1,
    max_size=20,
    echo=True
)

---

3. Modellarni Yaratish (Models)

Sizning modellariz bazadagi jadvallarni ifodalaydi. Sinxron modellar Model klassidan, asinxron modellar esa AsyncModel klassidan meros oladi.

from postgresdb3.orm.models import Model, AsyncModel
from postgresdb3 import String, Integer

# Sinxron model
class Category(Model):
    name = String(length=50, unique=True)

    class Meta:
        table_name = "categories"  # Jadval nomi (ixtiyoriy, berilmasa klass nomi olinadi)

# Asinxron model
class AsyncCategory(AsyncModel):
    name = String(length=50, unique=True)

Meta xususiyatlari:

• table_name yoki db_table (str): Bazadagi jadval nomi (masalan: table_name = "custom_table").
• abstract (bool): True bo'lsa, ushbu model uchun jadval yaratilmaydi (faqat boshqa modellarga meros qoldirish uchun xizmat qiladi).
• ordering (list/tuple): Standart saralash tartibi (masalan: ordering = ["-created_at", "id"]). Bu filtrlashda avtomatik ravishda ORDER BY created_at DESC, id ASC ko'rinishida qo'llanadi.
• unique_together (tuple): Bir nechta maydonlarning birgalikdagi takrorlanmasligi sharti.
• index_together (tuple): Birgalikda indeks yaratiladigan maydonlar guruhi.
• verbose_name / verbose_name_plural (str): Modelning inson tushunadigan tildagi yakka/ko'plik nomi.

Meta Merosxo'rligi (Meta Inheritance):

Ota model (Parent class) abstract bo'lsa, undan meros oluvchi subclasslar avtomatik tarzda ota klassning Meta xususiyatlarini (masalan, ordering, unique_together kabi) meros qilib oladi. Agar siz ota klass Meta xususiyatlarini saqlagan holda qo'shimcha qilmoqchi bo'lsangiz, Python'ning standart merosxo'rlik sintaksisidan foydalanishingiz mumkin:

class ParentModel(Model):
    class Meta:
        abstract = True
        ordering = ["-created_at"]

class ChildModel(ParentModel):
    class Meta(ParentModel.Meta):
        db_table = "custom_child_table" # Ota klassdagi ordering avtomatik saqlanib qoladi!

---

4. Maydonlar (Fields)

PostgresDB3 juda ko'p turdagi maydonlarni taqdim etadi. Har bir maydon ma'lum bir SQL turiga to'g'ri keladi.

Barcha maydonlar uchun umumiy parametrlar:

• verbose_name (str): Maydonning inson o'qishi uchun qulay bo'lgan nomi. Uni birinchi positional argument (masalan: String("To'liq ism", length=50)) yoki keyword argument (masalan: Integer(verbose_name="Yosh")) sifatida uzatish mumkin. Agar validation xatoligi yuzaga kelsa, ORM avtomatik ravishda ushbu nomdan foydalanib xatolik xabarini chiqaradi.
• nullable (bool): Maydon NULL qiymat qabul qiladimi? (Default: False)
• default (Any): Standart qiymat.
• primary_key (bool): Birlamchi kalitmi? (Default: False)
• unique (bool): Qiymat takrorlanmas bo'lishi kerakmi? (Default: False)
• validators (list): Maydon qiymatini tekshiruvchi chaqiriluvchi (callable) funksiyalar ro'yxati.

Matematik maydonlar:

• Integer: Oddiy butun son (integer).
• SmallInteger: Kichik butun son (smallint).
• BigInteger: Katta butun son (bigint).
• Float: Haqiqiy son (real).
• Double: Yuqori aniqlikdagi son (double precision).
• Decimal: Belgilangan aniqlikdagi o'nli kasr son.

Matnli maydonlar:

• String(length): Belgilangan uzunlikdagi satr (varchar). length parametri majburiy.
• Text: Cheklanmagan uzunlikdagi matn (text).

Mantiqiy va Maxsus maydonlar:

• Boolean: Mantiqiy maydon (boolean).
• UUID: UUID formatidagi maydon (uuid).
• JSON / JSONB: JSON formatidagi ma'lumotlar uchun.
• Array(item_type): Massivlar uchun (masalan: Array(Integer) -> integer[]).
• Serial / BigSerial: Avtomatik o'suvchi birlamchi kalitlar.

Aloqalar (Relationship Fields):

• ForeignKey(to, to_field=None, related_name=None, on_delete="CASCADE"): Ko'pga-bir (Many-to-One) aloqasi.
• OneToOneField(to, to_field=None, related_name=None, on_delete="CASCADE"): Birga-bir (One-to-One) aloqasi (avtomatik unique=True qo'shiladi).
• ManyToManyField(to, related_name=None): Ko'pga-ko'p (Many-to-Many) aloqasi (ikkala modelni bog'laydigan uchinchi jadval yaratiladi).

Aloqa parametrlari:

• to: Bog'lanayotgan maqsadli model klassi (target model).
• related_name (str): Qarama-qarshi modeldan ushbu modelga murojaat qilish uchun ishlatiladigan nom. Masalan, Post modelida author = ForeignKey(User, related_name="posts") deb yozilsa, user obyektidan uning barcha postlarini user.posts orqali olish imkoni yaratiladi.
• to_field (str): Maqsadli modeldagi bog'lanayotgan maydon nomi (sukut bo'yicha birlamchi kalit/primary key olinadi).
• on_delete: Bog'langan yozuv o'chirilganda bajariladigan amal ("CASCADE", "SET NULL", "RESTRICT", "NO ACTION"). Sukut bo'yicha: "CASCADE".

---

5. Validatorlar (Validation)

Ma'lumotlar bazaga saqlanishidan oldin qiymatlarni tekshirish uchun Django uslubidagi validatorlardan foydalaniladi. Tekshiruvdan o'tmagan holatda ValidationError xatoligi yuzaga keladi.

built-in validatorlar:

• MinValueValidator(limit_value, message=None): Qiymat belgilangan miqdordan kichik bo'lmasligini tekshiradi.
• MaxValueValidator(limit_value, message=None): Qiymat belgilangan miqdordan katta bo'lmasligini tekshiradi.
• MinLengthValidator(limit_value, message=None): Satr/Massiv uzunligi minimal shartga javob berishini tekshiradi.
• MaxLengthValidator(limit_value, message=None): Satr/Massiv uzunligi maksimal cheklovdan oshmasligini tekshiradi.
• RegexValidator(regex, message=None): Qiymat muntazam ifodaga (regular expression) mos kelishini tekshiradi.
• EmailValidator(message=None): Elektron pochta manzili formatini tekshiradi.

Ishlatilishi:

from postgresdb3 import String, Integer, EmailValidator, MinValueValidator

class User(Model):
    age = Integer(validators=[MinValueValidator(18)])
    email = String(length=100, validators=[EmailValidator()])

Maxsus validator (Custom Validator) yaratish:

Qiymat qabul qilib, xato bo'lsa ValidationError ko'taradigan oddiy funksiya yozish kifoya:

from postgresdb3 import ValidationError

def validate_even(value):
    if value % 2 != 0:
        raise ValidationError(f"{value} juft son bo'lishi shart!")

---

6. Model Metodlari va Hooklar (Hooks)

Siz model klassi metodlarini qayta yozib (override qilib), saqlash va o'chirish jarayonlarini nazorat qilishingiz mumkin.

1. clean() - Model darajasidagi tekshiruvlar

Bir nechta maydonlarni o'zaro solishtirish uchun ishlatiladi:

class User(Model):
    password = String(length=128)
    password_confirm = String(length=128)

    def clean(self):
        super().clean()  # Field validatorlarini ishga tushirish uchun!
        if self.password != self.password_confirm:
            raise ValidationError("Parollar o'zaro mos emas!")

2. before_save() va after_save(created: bool)

Ma'lumot yozilishidan oldin qiymatni o'zgartirish (masalan: parolni xeshlash) va yozilgandan keyin biron amal bajarish uchun:

import hashlib

class User(Model):
    username = String(length=50)
    password = String(length=128)
    password_hash = String(length=128, nullable=True)

    def before_save(self):
        if self.password:
            self.password_hash = hashlib.sha256(self.password.encode()).hexdigest()
            self.password = None  # Asl parolni tozalaymiz

    def after_save(self, created: bool):
        if created:
            print("Yangi foydalanuvchi yaratildi!")

3. before_delete() va after_delete()

Yozuv bazadan o'chirilishidan oldin va o'chirilgandan so'ng ishga tushadi.

---

7. CRUD Operatsiyalari

Sinxron va asinxron modellarda yozuvlarni yaratish, o'qish, yangilash va o'chirish usullari:

Yozuv Yaratish (Create)

# Sinxron:
user = User.create(username="ali", age=25)
# Yoki:
user = User(username="ali", age=25)
user.save()

# Asinxron:
user = await AsyncUser.create(username="ali", age=25)
# Yoki:
user = AsyncUser(username="ali", age=25)
await user.save()

Yozuv O'qish (Read)

# Sinxron:
user = User.query().filter(id=1).first()

# Asinxron:
user = await AsyncUser.query().filter(id=1).first()

Yozuv Yangilash (Update)

# Sinxron:
user.age = 26
user.save()

# Asinxron:
user.age = 26
await user.save()

Yozuv O'chirish (Delete)

# Sinxron:
user.delete()

# Asinxron:
await user.delete()

---

8. QuerySet API (Qidiruv va Filtrlash)

QuerySet orqali bazadan ma'lumotlarni turli shartlar bilan filtrlash, tartiblash va optimallashtirish amalga oshiriladi.

Asosiy qidiruv metodlari:

• .filter(*args, **kwargs): Shartga mos yozuvlarni olish.
• .exclude(*args, **kwargs): Shartga mos bo'lmagan yozuvlarni olish.
• .order_by(field): Tartiblash (teskari tartiblash uchun boshiga - qo'yiladi: -age).
• .limit(n): Natijalar sonini cheklash.
• .offset(n): Boshidan ma'lum miqdordagi yozuvlarni tashlab yuborish.
• .select_for_update(): Tranzaksiya doirasida tanlangan yozuvlarni qulflash (Pessimistic locking). Balanslarni yangilash, to'lovlar yoki zaxira mahsulotlarini kamaytirish kabi parallel poyga holatlarining (race condition) oldini oladi.

Misollar:

# Yosh 18 dan katta va ism 'A' harfi bilan boshlanadigan foydalanuvchilar
users = User.query().filter(age__gt=18, name__startswith="A").order_by("-age").all()

Qidiruv shablonlari (Lookup field suffixes):

• __gt / __gte: Katta / Katta yoki teng.
• __lt / __lte: Kichik / Kichik yoki teng.
• __contains / __icontains: Matn ichida mavjudligi (registrsiz).
• __startswith / __endswith: Matn boshlanishi / tugashi.
• __in: Berilgan ro'yxat ichida mavjudligi (id__in=[1, 2, 3]).

Murakkab Shartlar (Q va F Expressions):

Q - shartlarni OR (|) yoki AND (&) orqali bog'lash uchun. F - maydon qiymatini boshqa maydon bilan solishtirish uchun.

from postgresdb3 import Q, F

# Ismi 'Ali' YOKI yoshi 20 dan katta bo'lganlar
users = User.query().filter(Q(name="Ali") | Q(age__gt=20)).all()

# Ball yoshidan baland bo'lganlar
users = User.query().filter(score__gt=F("age")).all()

Ommaviy Operatsiyalar (Bulk Operations)

1. Query-level ommaviy yangilash va o'chirish:

Bazada birdaniga bir nechta qatorlarni Python xotirasiga yuklamasdan to'g'ridan-to'g'ri o'zgartirish yoki o'chirish (avtomatik himoyalangan: shart kiritilishi majburiy):

# Yosh 18 dan kichik bo'lgan barcha foydalanuvchilar ballini 0 ga tushirish
User.query().filter(age__lt=18).update(score=0.0)

# Balli 0 bo'lgan barcha foydalanuvchilarni o'chirish
User.query().filter(score=0.0).delete()

2. bulk_create (Klass darajasidagi ommaviy yaratish):

Ko'plab model nusxalarini (instances) bitta INSERT so'rovi orqali bazaga juda tez yozish uchun ishlatiladi:

users = [
    User(name="User 1", age=20),
    User(name="User 2", age=25),
    User(name="User 3", age=30),
]
# Sinxron:
User.bulk_create(users)

# Asinxron:
await AsyncUser.bulk_create(async_users)

3. bulk_update (Klass darajasidagi ommaviy yangilash):

Mavjud model nusxalarining belgilangan maydonlarini bitta so'rov orqali bazada yangilash uchun xizmat qiladi:

# Model obyektlarining maydonlarini o'zgartiramiz
for user in users_list:
    user.age += 1

# Sinxron (faqat belgilangan 'age' maydoni bazada yangilanadi):
User.bulk_update(users_list, fields=["age"])

# Asinxron:
await AsyncUser.bulk_update(async_users_list, fields=["age"])

---

Tranzaksiyalar (Transactions)

PostgresDB3 tranzaksiyalar bilan ishlashni juda qulay va xavfsiz qiladi. Buning uchun ikki xil usul mavjud:

1. Context Manager sifatida (with / async with):

Kodni ma'lum bir qismini tranzaksiya ichida bajarish uchun:

# Sinxron:
with User.db.transaction():
    user1 = User.create(name="Ali", age=22)
    user2 = User.create(name="Vali", age=25)

# Asinxron:
async with AsyncUser.db.transaction():
    await AsyncUser.create(name="Ali", age=22)
    await AsyncUser.create(name="Vali", age=25)

2. Decorator sifatida (@db.atomic()):

Butun funksiyani avtomatik tranzaksiyaga o'rab qo'yish uchun:

# Sinxron:
@User.db.atomic()
def transfer_funds(sender_id, receiver_id, amount):
    sender = User.query().select_for_update().filter(id=sender_id).first()
    receiver = User.query().select_for_update().filter(id=receiver_id).first()
    
    sender.balance -= amount
    receiver.balance += amount
    sender.save()
    receiver.save()

# Asinxron:
@AsyncUser.db.atomic()
async def async_transfer_funds(sender_id, receiver_id, amount):
    sender = await AsyncUser.query().select_for_update().filter(id=sender_id).first()
    receiver = await AsyncUser.query().select_for_update().filter(id=receiver_id).first()
    
    sender.balance -= amount
    receiver.balance += amount
    await sender.save()
    await receiver.save()

---

Sahifalash (Pagination)

# Sinxron (PaginationResult obyekti qaytadi):
result = User.query().paginate(page=1, per_page=10)
print(result.total)        # Jami yozuvlar soni
print(result.pages)        # Jami sahifalar
print(result.data)         # Model nusxalari ro'yxati (List[User])

# Asinxron (Lug'at ko'rinishida qaytadi):
result = await AsyncUser.query().paginate(page=1, per_page=10)
print(result["total"])
print(result["data"])      # List[AsyncUser]

Aggregatsiyalar va Hisob-kitoblar (aggregate & annotate):

• Count, Sum, Avg, Max, Min funktsiyalarini qo'llash:

from postgresdb3 import Sum, Avg, Count

# O'rtacha yosh va jami ballni hisoblash
stats = User.query().aggregate(avg_age=Avg("age"), total_score=Sum("score"))
print(stats)  # {'avg_age': 25.4, 'total_score': 1250.0}

# Har bir foydalanuvchi postlari sonini qo'shib olish
users = User.query().annotate(posts_count=Count("posts")).all()
print(users[0].posts_count)

Bog'lanishlarni yuklash (N+1 muammosini hal qilish):

• Sinxron / Asinxron aloqalarni optimallashtirish:

# ForeignKey uchun JOIN ishlatadi (select_related)
posts = Post.query().select_related("author").all()

# ManyToMany yoki OneToMany uchun alohida so'rov bilan yig'adi (prefetch_related)
posts = Post.query().prefetch_related("tags").all()

---

Sof SQL So'rovlari (Raw SQL Queries)

Murakkab so'rovlar yoki to'g'ridan-to'g'ri SQL kodini yozish kerak bo'lgan holatlar uchun PostgresDB3 quyidagi imkoniyatlarni taqdim etadi:

1. Model nusxalarini qaytaruvchi Raw SQL (raw_sql):

Sof SQL yozib, natijalarni avtomatik ravishda tegishli Model obyektlari (instances) ko'rinishida olish uchun:

# Sinxron (psycopg2 parametrlaridan foydalanadi: %s):
users = User.raw_sql("SELECT * FROM users WHERE age > %s AND status = %s", 18, "active")
for user in users:
    print(user.name)  # Obyekt maydonlariga to'g'ridan-to'g'ri murojaat qilish

# Asinxron (asyncpg parametrlari: $1, $2):
users = await AsyncUser.raw_sql("SELECT * FROM users WHERE age > $1 AND status = $2", 18, "active")

2. Bazaning pastki darajali drayveri orqali Raw SQL:

Model obyektlarisiz, oddiy lug'at (dictionary) yoki ro'yxat ko'rinishidagi ma'lumotlarni to'g'ridan-to'g'ri olish yoki bazaga o'zgartirish kiritish uchun:

# Sinxron (db.raw orqali):
record = User.db.raw("SELECT name, balance FROM users WHERE id = %s", [1], fetchone=True)
print(record["name"])

# Asinxron (db._manager orqali):
records = await AsyncUser.db._manager("SELECT name, balance FROM users WHERE age > $1", 18, fetchall=True)

---

9. Aloqalar bilan ishlash (Relationships)

PostgresDB3 munosabatlarni to'g'ri o'rnatish, ma'lumotlarni bog'lash, so'rovlarni avtomatik optimallashtirish va Django-style qulayliklarni qo'llab-quvvatlaydi.

A) Sinxron modellarda aloqalar bilan ishlash (Model)

1. ForeignKey (One-to-Many) va OneToOneField (Birga-bir)

Kutubxona maydon nomini avtomatik tarzda database ustuniga (_id qo'shimchasi bilan) xaritlaydi va descriptorlar orqali aloqalarni boshqaradi.

# Modellar:
class User(Model):
    table = "users"
    name = String()

class Order(Model):
    table = "orders"
    user = ForeignKey(User)  # Bazada 'user_id' ustuni yaratiladi
    amount = Integer()

# Ma'lumot qo'shish (Django-style, obyektning o'zini uzatish):
new_user = User.create(name="Ali")
order = Order.create(user=new_user, amount=50000)

# Yoki ID orqali saqlash:
order = Order.create(user_id=new_user.id, amount=50000)

# Bog'langan model ma'lumotini o'qish (Lazy Loading):
fetched_order = Order.filter(id=order.id).first()
related_user = fetched_order.user  # Bazadan avtomatik yuklanadi
print(related_user.name)  # Ali

2. ManyToManyField (Ko'pga-ko'p)

ManyToMany munosabatlarida bog'liqliklar uchinchi oraliq jadvalda ({table1}_{table2}) saqlanadi va ularni boshqarish uchun .add(), .remove(), .clear() va .all() metodlaridan foydalaniladi.

# Modellar:
class Tag(Model):
    table = "tags"
    name = String()

class Product(Model):
    table = "products"
    tags = ManyToManyField(Tag)
    price = Integer()

product = Product.create(price=15000)
tag_new = Tag.create(name="Yangi")
tag_sale = Tag.create(name="Chegirma")

# Bog'lash (Obyektlar yoki ID-lar orqali):
product.tags.add(tag_new, tag_sale)

# Bog'liqlikni o'chirish:
product.tags.remove(tag_new)

# Barcha bog'langan teglarni o'qish:
tags = product.tags.all()

# Barcha bog'liqliklarni tozalash:
product.tags.clear()

---

B) Asinxron modellarda aloqalar bilan ishlash (AsyncModel)

Asinxron rejimda aloqalar va metodlar await orqali ishlatilishi lozim.

1. ForeignKey va Lazy Loading

# Modellar:
class AsyncUser(AsyncModel):
    table = "users"
    name = String()

class AsyncOrder(AsyncModel):
    table = "orders"
    user = ForeignKey(AsyncUser)
    amount = Integer()

# Asinxron yozuv yaratish:
user = await AsyncUser.create(name="Ali (Async)")
order = await AsyncOrder.create(user=user, amount=120000)

# Asinxron Lazy Loading (e'tibor bering, bog'lanish await qilinadi):
fetched_order = await AsyncOrder.filter(id=order.id).first()
related_user = await fetched_order.user  # COROUTINE obyekti await qilinadi
print(related_user.name)

2. ManyToManyField

# Asinxron bog'lanish qo'shish:
await product.tags.add(tag1, tag2)

# Asinxron bog'lanish o'chirish:
await product.tags.remove(tag1)

# Asinxron barcha bog'liqliklarni o'qib olish:
tags = await product.tags.all()

# Asinxron tozalash:
await product.tags.clear()

---

10. Migratsiya Tizimi (CLI va Kod orqali)

PostgresDB3 o'z ichida model o'zgarishlarini kuzatib boruvchi va bazadagi jadvallarni avtomatik yangilovchi migratsiya dvigateliga ega.

A) Terminal orqali boshqarish (CLI)

Avvalo loyihangizda manage.py faylini yarating:

import sys
from postgresdb3 import execute_from_command_line
from myapp.models import db_sync  # Ulanish obyekti
from myapp.models import User, Post  # Barcha modellar import qilinishi shart!

if __name__ == "__main__":
    execute_from_command_line(db_sync, sys.argv)

Terminal komandalari:

# 1. Modellarni tahlil qilib migratsiya faylini yaratish
python manage.py makemigrations initial_setup

# 1.1. CI/CD pipelines yoki avtomatlashtirilgan (Docker) tizimlarda so'rovlarsiz ishga tushirish uchun:
python manage.py makemigrations initial_setup --no-input

# 2. Yaratilgan migratsiyalarni bazaga qo'llash (jadvallarni yaratish)
python manage.py migrate

# 3. Oxirgi migratsiyani bekor qilish (Rollback)
python manage.py undo

B) Dasturiy ravishda boshqarish (Programmatic)

Terminal bo'lmagan muhitlarda (masalan: dastur kodining o'zida) migratsiyalarni bajarish:

from postgresdb3.migrations.engine import MigrationEngine
from myapp.models import db_sync, db_async

engine = MigrationEngine()

# 1. Migratsiya faylini tayyorlash
# (interactive=False bo'lsa, ogohlantirish so'ramaydi - CI/CD uchun qulay)
engine.makemigrations(name="auto_setup", interactive=False)

# 2. Sinxron bazani yangilash
engine.migrate(db_sync)

# 3. Asinxron bazani yangilash (asinxron funksiya ichida chaqiriladi)
await engine.async_migrate(db_async)

# 4. Sinxron/Asinxron migratsiyani orqaga qaytarish
engine.undo_migration(db_sync)
await engine.async_undo_migration(db_async)

---

📄 Litsenziya

Ushbu loyiha MIT litsenziyasi ostida tarqatiladi.