Dep injector

Allows expressing dependency graphs and getting dependencies from them.

This is potentially useful for dependency injection.

Module docs:

• injector package
  • injector.dependencies module
  • injector.exceptions module
  • injector.graph module
  • injector.injector module

Example:

from injector import Dependencies

def value_source():
    return 'Result of a {}'.format('computation')

def both_source_and_dependant(computed_value, another_value):
    return len(computed_value) + another_value

class UsesInjector:
    def __init__(self, computed_value, len_value):
        self.s = computed_value
        self.l = len_value

    def __str__(self):
        return 'UsesInjector({}, {})'.format(self.s, self.l)

# register dependencies
deps = Dependencies()
# You can register a class as a factory
deps.register_factory(
    'uses-injector',
    UsesInjector,
    dependencies=['computed_value', 'len_value'],
)
# or a function
deps.register_factory(
    'len_value',
    both_source_and_dependant,
    dependencies=['computed_value', 'another_value'],
)
deps.register_factory('computed_value', value_source)
# You can also register plain values
deps.register_value('another_value', 10000)

# Get the injector - this finalizes the dependencies and checks to make sure
# that the dependencies are all correct.
inj = deps.build_injector()

print(inj.get_dependency('uses-injector'))
print(inj.get_dependency('len_value'))

# You can also inject new functions (just not register more values)
def uses_computed_value(computed_value):
    return ''.join(set(computed_value))

print(inj.inject(uses_computed_value, ['computed_value']))

Development

Run this:

virtualenv -p python3 env
source env/bin/activate
pip install -r dev-requirements.txt
nosetests .