https://github.com/hydrogen602/easy-serialize
Easily convert custom python objects to string and back
https://github.com/hydrogen602/easy-serialize
json python serialization
Last synced: about 2 months ago
JSON representation
Easily convert custom python objects to string and back
- Host: GitHub
- URL: https://github.com/hydrogen602/easy-serialize
- Owner: hydrogen602
- License: mit
- Created: 2021-04-28T18:07:32.000Z (about 5 years ago)
- Default Branch: main
- Last Pushed: 2021-10-10T21:46:55.000Z (over 4 years ago)
- Last Synced: 2025-02-17T18:03:48.727Z (over 1 year ago)
- Topics: json, python, serialization
- Language: Python
- Homepage: https://pypi.org/project/easy-serialize/
- Size: 22.5 KB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: readme.md
- License: LICENSE
Awesome Lists containing this project
README
# Easy-Serialize
Turn custom python objects into strings and vice-versa.
## Strengths
- no extra code to write in most cases
- supports json
- supports nested objects
- no dependencies outside of python's library (only pytest for developing)
- ability to add methods to control serializing and deserializing
- type hints and works with mypy
## Drawbacks
- objects like tuples are automatically turned into lists
- only supports json right now
- can't handle circular references
- if an object has references in many places, it will be present in the json repeatedly and after deserializing will be separate objects
- Because of this, it cannot handle linked lists or similar interconnected data structures well
- stores `__dict__` of the object, so classes using `__slots__` will likely require overriding the `serialize` method and `deserialize` classmethod.
## How to use
Extend `Serializable`
```
class A(Serializable):
...
```
or use the `make_serializable` decorator
```
@make_serializable
class A:
...
```
serialize using
```
serialize(object_to_serialize)
```
and deserialize using
```
deserialize(stringified_object)
```
### Example
Refer to [Simple Example](demo.py)
### Custom serializing
In the case that the standard method doesn't cut it, it is
possible to add custom serializing and deserializing methods to a class
```
@make_serializable
class A:
def __init__(self, x):
self.x = x
def serialize(self) -> dict:
return {'x': self.x}
@classmethod
def deserialize(cls, data: dict) -> 'A':
return A(data['x'])
```
## Notes
- `__init__` is by default not called for deserialized objects
- Because it needs to create the object and then copy over all the values of `__dict__`, it doesn't know the arguments used in `__init__`. Thus the object will be created with `__new__` which avoids calling `__init__`.
## Future
Things to possibly add:
- other data formats like xml or the like
- support for objects with circular references