Slowstore

Simple key-value store that stores your objects as JSON on file system.

It directly maps your python objects to JSON files and every object is auto synced to its JSON file.

It is designed to be easy to plug into your program, without need of servers, connection strings, or any setup just set the directory where you would like to store your files.

*Slowstore it is intended to be used for exploration and debugging of my program behaviour.

Installation

You can install Slowstore using pip:

pip install slowstore

Usage

You can create a new Slowstore instance by calling the constructor:

from slowstore import Slowstore
from pydantic import BaseModel

class SampleModel(BaseModel):
    name: str
    age: int = 0

    def birthday(self):
        self.age += 1

# Create the store to save data under "mydata" directory
store = Slowstore[SampleModel](SampleModel, "mydata")

# This is how you add or update an object in the store
dennis = store.upsert("dennis", SampleModel(name="denis", age=32))

# immediately after previous line is evaluated,
# you will have a json file (mydata/dennis.json) represening this object

dennis.name = "DENIS"
# here the name in the json will also change from "dennis" to "DENIS"
# also the associated change will be tracked so you can further inspect if needed.

dennis.birthday()
# will also trigger another change in the age field and it will be reflected in the json file. 

mydata/dennis.json after running this program will look like:

{
  "name": "DENIS",
  "age": 33,
  "__key__": "dennis",
  "__changes__": [
    {
      "key": "dennis",
      "prop_name": "age",
      "prev_val": 32,
      "new_val": 33,
      "date": "2024-08-28T19:04:12.840353"
    },
    {
      "key": "dennis",
      "prop_name": "name",
      "prev_val": "denis",
      "new_val": "DENIS",
      "date": "2024-08-28T19:04:12.840216"
    },
  ]
}

Slowstore tracks what happened in your small program on every field. You can toggle the save_on_change flat at any time after the store is created, or you can also set the save_on_change flag to your liking in the constructor of the store

store.save_on_change = False 

After this code is executed, you will need to run store.commit(some_model) or store.commit_all() in order to persist the changes.

Commit the changes to the store

If the store's save_on_change flag is not set to True, you can commit the changes manually.

store.commit(s1)
store.commit(s2)
# or just commit all changes
store.commit_all() 

Undo/Redo

Every slowstore item is a proxy object around the object you added to the store. This allows you to access the object's attributes and adds undo/redo/dirty functionality

s1.undo()
s1.redo()

Querying the store

# get all items, returns an iterator
all_items = store.all()

# yield all items that match a condition
filtered_items = store.filter(lambda x: x.age > 30)

# get the first item that matches a condition
first_item = store.first(lambda x: x.age > 30)

Check if some object or it's key is in the store

if s2 in store:
    print("s2 is in store")
else:
    print("s2 is not in store")

Deleting items

Any of the following will delete the item from the store, along with the file that contains the data

store.delete(s1)
del store[s2]
key = "some_key"
store.delete(key)

How is the data organized?

The data is stored in a directory, with each item stored in a separate file. The file name is the key of the item, and the content is the JSON representation of the object.

That's all there is to it.

Features

At the moment Slowstore works with Pydantic models, but I plan to add support for other types of objects that are serializable to JSON.

• Save on change: Set the store to save the data every time a property is changed
• Undo/Redo: Undo and redo changes to the object
• Dirty: Check if the object has been changed
• Filtering: Filter/Query the items in the store
• Deleting: Delete items from the store
• Commit: Commit changes to the store
• Partial load: Load only the items you need, and lazy load the rest
• Transactions: Add support for transactions
• Non-Pydantic objects: Add support for other types of objects
• Indexes: Add indexes to the store to speed up queries

How it works

The slowtore instance behaves like a dictionary of objects you added to it. Instead of storing the object it creates a proxy object that stores the object's state and the changes you made to it.

When you commit the changes, the proxy object is updated with the new state, and the changes are saved to disk.

When you undo/redo the changes, the proxy object is updated with the previous state, and the changes are saved to disk.

Check the ModelProxy class for more details.

Important

Slow store is slow and it is not intended to be used in multithreaded contexts. It's primarily created to provide a good DX while working on a certain feature.

License

This project is licensed under the MIT License - see the LICENSE file for details.