flashback 1.2.0

An utility library for python

Flashback

You've probably already implemented it.

---

Flashback is a collection of python utilities: it contains classes, metaclasses, decorators, functions, constants, etc. that you might have written a couple times already.

Installing

Flashback supports python 3.6+.

Pip:

pip install flashback

Build from sources:

git clone [e-mail]:PaulRenvoise/flashback.git
cd flashback
python setup.py install

Contents

Flashback's helpers are currently organised within 7 modules, and global helpers:

• accessing/
  • dig() recursively fetch keys in a nested dict
• caching/
  • Cache supports several cache stores: in-memory, disk, Redis, and Memcached
  • @cached caches a callable's return value based on its arguments
• debugging/
  • xp() prints debug information about its given arguments
  • @profiled collects and dumps profiling stats over a callable's execution
  • caller() allows a developer to print debug information about a callable's caller
  • get_callable() extracts a callable instance from a frame
  • get_call_context() finds and returning the code context around a call made in a frame
  • get_frameinfo() implements a faster inspect.stack()[x]
• formatting/
  • oxford_join() joins strings in a human-readable way
  • transliterate() represents unicode text in ASCII (using Unidecode)
  • camelize() transforms any case to camelCase
  • pascalize() transforms any case to PascalCase
  • snakeize() transforms any case to snake_case
  • kebabize() transforms any case to kebab-case
  • parameterize() formats a given string to be used in URLs
  • ordinalize() represents numbers in their ordinal representations
  • adverbize() represents numbers in their numeral adverb representations
  • truncate() truncates long sentences at a given limit and append a suffix if needed
  • singularize() returns the singular form of a given word
  • pluralize() returns the plural form of a given word
• iterating/
  • renumerate() enumerates an iterable starting from its end
  • chunks() splits an iterable into smaller chunks, padding them if requested
  • partition() splits an iterable into items validating a predicate and the ones that don't
  • uniq() removes duplicates from an iterable while keeping the items' order
  • compact() removes None values from an iterable
  • flatten() unpacks nested iterable into the given iterable
  • flat_map() applies a function to every item and nested item of the given iterable
• i16g/
  • Locale dynamically loads localization files from a package path
• importing/
  • import_class_from_path() fetches a class from a package path and returns it
  • import_module_from_path() exposes the contents of a module as globals from a package path
• logging/
  • DEFAULT_CONSOLE_CONFIGURATION logs to stderr with a sensible set of information
  • DJANGO_CONSOLE_CONFIGURATION logs to stderr with the same formatting as Django's logger
  • FLASK_CONSOLE_CONFIGURATION logs to stderr with the same formatting as Flask's logger
  • PYRAMID_CONSOLE_CONFIGURATION logs to stderr with the same formatting as Pyramid's logger
  • RAILS_CONSOLE_CONFIGURATION logs to stderr with the same formatting as RoR's logger
  • AffixedStreamHandler allows custom affixes to log records
  • @muted silences all (or selected) loggers during a callable's execution
• Borg exposes a class useful to produce a singleton behaviour across multiple instances
• Sentinel exposes a class that can be used to implement the Sentinel design pattern
• Singleton exposes a metaclass useful to implement the Singleton design pattern
• @classproperty combines @classmethod and @property (with support for @attr.setter)
• @deprecated documents deprecated callables with a explicit message
• @retryable retries failing executing of a callable
• @sampled implements sampling strategies to filter calls made to a callable
• @timed measures and prints the execution time of a callable
• @timeoutable stops the execution of a callable if its run time is too long

Contributing

The Pull Request template has a checklist containing everything you need to submit a new PR.

Run the tests with pytest:

pytest tests

Run the lint with pylint:

pylint flashback/ tests/

License

Flashback is released under the MIT License.