class-cache 0.10.1

Caching for class based generators

Class based cache

Installation

Recommended installation with pip:

pip install class-cache

Usage

• Basic usage:

  from class_cache import Cache
  
  # Create cache
  cache = Cache()
  # Set item in cache
  # NOTE: Keys and values have to be pickle-serialisable
  cache["foo"] = "bar"
  # Save cache to backend (disk by default)
  cache.write()
  
  # During another program run just create same cache again and you can retrieve data
  del cache
  cache2 = Cache()
  assert cache2["foo"] == "bar"
  

• Use multiple caches:

  cache1 = Cache(1)
  cache2 = Cache(2)
  
  cache1["foo"] = "bar"
  cache2["foo"] = "zar"
  
  assert cache1["foo"] != cache2["foo"]
  

• Use cache with default factory:

  from class_cache import CacheWithDefault
  
  class MyCache(CacheWithDefault[str, str]):
      NON_HASH_ATTRIBUTES = frozenset({*CacheWithDefault.NON_HASH_ATTRIBUTES, "_misc"})
      def __init__(self, name: str):
          # Attributes which affect default value generation should come before super().__init__()
          # They will be used to generate a unique id
          self._name = name
          super().__init__()
          # Other attributes should not affect how default value is generated, add them to NON_HASH_ATTRIBUTES
          self._misc = "foo"
  
      # Define logic for defaults in _get_data
      def _get_data(self, key: str) -> str:
          return f"{self._name}_{key}"
  
  cache = MyCache("first")
  assert cache["foo"] == "first_foo"
  

• Compress data before storing:

  from class_cache.wrappers import BrotliCompressWrapper
  from class_cache import Cache
  
  cache = BrotliCompressWrapper(Cache())
  # Use cache as normal
  

  This wrapper uses brotli algorithm for compression, which optimises read-time at expense of write-time. This will generally lead to less space being used and potentially faster reads if your data is compressible, e.g. text.

Development

• Install dev dependencies: pip install -e ".[dev]"

• For linting and basic fixes ruff is used: ruff check . --fix

• This repository follows strict formatting style which will be checked by the CI.

  • To format the code, use the black format: black .
  • To sort the imports, user isort utility: isort .

• To test code, use pytest: pytest .

• This repository follows semantic-release, which means all commit messages have to follow a style. You can use tools like commitizen to write your commits.

• You can also use pre-commit to help verify that all changes are valid. Multiple hooks are used, so use the following commands to install:

  pre-commit install
  pre-commit install --hook-type commit-msg