Generate two lists of equal length in fixed_dictionaries using Hypothesis

Question

I'm trying to generate sample data using the fixed_dictionaries strategy where two of the keys have lists as values that must be of equal length, e.g.:

{'ids': [1, 2, 3],
 'words': ['foo', 'bar', 'baz']}

How can I enforce this constraint? I thought I might be able to define one as a composite in terms of the other, but I'm not sure how to do that. Something like:

import hypothesis.strategies as st

ids = st.lists(elements=st.integers())

@st.composite
def words(draw, elements=st.text()):
    draw(sample_ids) # ???

why not use a single key for both, and a list of id,word tuples as data? — Jean-François Fabre, Sep 21 '18 at 15:56
Because that's not how the data I'm trying to test is shaped. — tkocmathla, Sep 21 '18 at 15:57
when you say "two of the keys are lists" do you mean values? — D_________, Sep 21 '18 at 15:59
Suppose in a given case you get misshapen data, do you just want to drop the extra elements? Or do you want to throw an `Exception`? — C.Nivs, Sep 21 '18 at 16:06

score 6 · Accepted Answer · answered Sep 21 '18 at 16:15

Here's one way to solve this:

@composite
def my_dicts(draw):
    size = draw(st.integers(min_value=0))
    ids = st.lists(min_size=size, max_size=size)
    words = st.lists(min_size=size, max_size=size)
    dicts = st.fixed_dictionaries({'ids': ids, 'words': words})

    return draw(st.lists(dicts))

Generate two lists of equal length in fixed_dictionaries using Hypothesis

1 Answers1