The module provides extra resources to users of eth-abi that are not required for typical use. It can be installed with pip as an extra requirement:

pip install eth-abi[tools]

ABI Type Strategies

The tools module provides the get_abi_strategy() function. This function returns a hypothesis strategy (value generator) for any given ABI type specified by its canonical string representation:

>>> from import get_abi_strategy
>>> import random

>>> uint_st = get_abi_strategy('uint8')
>>> uint_st
integers(min_value=0, max_value=255)
>>> uint_st.example(random.Random(0))

>>> uint_list_st = get_abi_strategy('uint8[2]')
>>> uint_list_st
lists(elements=integers(min_value=0, max_value=255), min_size=2, max_size=2)
>>> uint_list_st.example(random.Random(0))
[66, 247]

>>> fixed_st = get_abi_strategy('fixed8x1')
>>> fixed_st
decimals(min_value=-128, max_value=127, places=0).map(scale_by_Eneg1)
>>> fixed_st.example(random.Random(0))

>>> tuple_st = get_abi_strategy('(bool,string)')
>>> tuple_st
tuples(booleans(), text())
>>> tuple_st.example(random.Random(0))
(False, '')


In the above code snippet, we use the example method on hypothesis strategy objects with an explicit random number generator with seed zero. Both the use of that method and the use of zero-seeded random number generators are not recommended for normal use. Their use in this context is intended to make the code snippet reproducible. In normal test code, use of the example method may prevent hypothesis from finding a useful sample of values. See here for more information.

Hypothesis strategies can be used to conduct property testing on contract code. For more information on property testing, visit the Hypothesis homepage or the Hypothesis readthedocs site.