Dynamesa

A Simple DynamoDB client

Makes the simplest operations simple, and smoothes over boto3’s quirks.

Installing Dynamesa

For now install from the repo. If boto3 isn’t already available in your python environment, you’ll need that too.

pip install -U git+https://github.com/jiaaro/dynamesa.git#egg=dynamesa

Finding tables

import dynamesa

# boto3 resource kwargs
dynamesa.configure(region_name="us-east-1")

# Finds tables automatically
table = dynamesa.tables.MyAppUsers

# allows for weird table names
table = dynamesa.tables["My-App-Users"]

# Get Tables with a dataclass to get back items as instances instead of dictionaries 
table = dynamesa.tables.get("My-App-Users", MyDataClass)

# You can also instantiate a table yourself (requires it's own configuration)
table = dynamesa.Table("myapp-users", region_name="us-east-1")

Example use

import dynamesa

table = dynamesa.tables.create(
    "User",
    ("id", "S", "ts", "N"),
    gsis={"AgeIndex": ("age", "N")},
)

table.put({"id": 1, "name": "Jack Frost", "age": "I'll never tell"})
table.get(id=1)

updated_item = table.update({
  "id": 1,
  "email": "jfrost@northpole.io",
  "age": dynamesa.REMOVE_KEY
})

# updated_item == {"id": 1, "name": "Jack Frost", "email": "jfrost@northpole.io"}

# returns iterator, uses an index, name isn't in the index so also uses a filter expression
table.find("UserAgeIndex", age="74", name="King Cole")

# Use the primary key index
table.find(dynamesa.PRIMARY_KEY, id=1)

# Query or scan using boto3 Key expressions 
table.find(dynamesa.PRIMARY_KEY, dynamesa.Key("id").eq(1))

# will scan (returns iterator)
table.find(email="jfrost@northpole.io")

# delete users who are 74 years old (using the age index)
table.clear("UserAgeIndex", age=74)

# delete everything
table.clear()

Example with Typed items

import dataclasses
import time
import dynamesa


@dataclasses.dataclass
class UserModel:
    id: str
    name: str
    email: str
    age: int = None

dynamesa.tables.create(
    "User",
    ("id", "S", "ts", "N"),
    gsis={"AgeIndex": ("age", "N")},
)
table = dynamesa.tables.get('User', UserModel)
jack = UserModel(1, "Jack Frost", "jfrost@northpole.net")
table.put(jack)

user = table.get(id=1)
user.name == "Jack Frost"