How to use compat_kw method in hypothesis

Best Python code snippet using hypothesis

numpy.py

Source:numpy.py

...65    # Subarray datatypes, eg '(2, 3)i4'66    if dtype.subdtype is not None:67        subtype, shape = dtype.subdtype68        return arrays(subtype, shape, elements=kwargs)69    def compat_kw(*args, **kw):70        """Update default args to the strategy with user-supplied keyword args."""71        assert {"min_value", "max_value", "max_size"}.issuperset(kw)72        for key in set(kwargs).intersection(kw):73            msg = f"dtype {dtype!r} requires {key}={kwargs[key]!r} to be %s {kw[key]!r}"74            if kw[key] is not None:75                if key.startswith("min_") and kw[key] > kwargs[key]:76                    raise InvalidArgument(msg % ("at least",))77                elif key.startswith("max_") and kw[key] < kwargs[key]:78                    raise InvalidArgument(msg % ("at most",))79        kw.update({k: v for k, v in kwargs.items() if k in args or k in kw})80        return kw81    # Scalar datatypes82    if dtype.kind == "b":83        result = st.booleans()  # type: SearchStrategy[Any]84    elif dtype.kind == "f":85        result = st.floats(86            width=8 * dtype.itemsize,87            **compat_kw(88                "min_value",89                "max_value",90                "allow_nan",91                "allow_infinity",92                "exclude_min",93                "exclude_max",94            ),95        )96    elif dtype.kind == "c":97        # If anyone wants to add a `width` argument to `complex_numbers()`, we would98        # accept a pull request and add passthrough support for magnitude bounds,99        # but it's a low priority otherwise.100        if dtype.itemsize == 8:101            float32 = st.floats(width=32, **compat_kw("allow_nan", "allow_infinity"))102            result = st.builds(complex, float32, float32)103        else:104            result = st.complex_numbers(**compat_kw("allow_nan", "allow_infinity"))105    elif dtype.kind in ("S", "a"):106        # Numpy strings are null-terminated; only allow round-trippable values.107        # `itemsize == 0` means 'fixed length determined at array creation'108        max_size = dtype.itemsize or None109        result = st.binary(**compat_kw("min_size", max_size=max_size)).filter(110            lambda b: b[-1:] != b"\0"111        )112    elif dtype.kind == "u":113        kw = compat_kw(min_value=0, max_value=2 ** (8 * dtype.itemsize) - 1)114        result = st.integers(**kw)115    elif dtype.kind == "i":116        overflow = 2 ** (8 * dtype.itemsize - 1)117        result = st.integers(**compat_kw(min_value=-overflow, max_value=overflow - 1))118    elif dtype.kind == "U":119        # Encoded in UTF-32 (four bytes/codepoint) and null-terminated120        max_size = (dtype.itemsize or 0) // 4 or None121        result = st.text(**compat_kw("alphabet", "min_size", max_size=max_size)).filter(122            lambda b: b[-1:] != "\0"123        )124    elif dtype.kind in ("m", "M"):125        if "[" in dtype.str:126            res = st.just(dtype.str.split("[")[-1][:-1])127        else:128            # Note that this case isn't valid to pass to arrays(), but we support129            # it here because we'd have to guard against equivalents in arrays()130            # regardless and drawing scalars is a valid use-case.131            res = st.sampled_from(TIME_RESOLUTIONS)132        result = st.builds(dtype.type, st.integers(-(2 ** 63), 2 ** 63 - 1), res)133    else:134        raise InvalidArgument(f"No strategy inference for {dtype}")135    return result.map(dtype.type)...

Automation Testing Tutorials

Learn to execute automation testing from scratch with LambdaTest Learning Hub. Right from setting up the prerequisites to run your first automation test, to following best practices and diving deeper into advanced test scenarios. LambdaTest Learning Hubs compile a list of step-by-step guides to help you be proficient with different test automation frameworks i.e. Selenium, Cypress, TestNG etc.