Merge pull request #3414 from Zac-HD/integer-boundaries

Generate integer bounds more often

hypothesis-python/RELEASE.rst

@@ -0,0 +1,4 @@
+RELEASE_TYPE: patch
+
+This patch makes :func:`~hypothesis.strategies.integers` more likely to
+generate boundary values for large two-sided intervals (:issue:`2942`).

hypothesis-python/scripts/basic-test.sh

@@ -81,6 +81,7 @@ if [ "$(python -c 'import platform; print(platform.python_implementation())')" !
   pip uninstall -y django pytz
 
   pip install "$(grep 'numpy==' ../requirements/coverage.txt)"
+  $PYTEST tests/array_api
   $PYTEST tests/numpy
 
   pip install "$(grep 'pandas==' ../requirements/coverage.txt)"

hypothesis-python/src/hypothesis/internal/conjecture/utils.py

@@ -63,9 +63,14 @@ def unbounded_integers(data: "ConjectureData") -> int:
 
 
 def integer_range(
-    data: "ConjectureData", lower: int, upper: int, center: Optional[int] = None
+    data: "ConjectureData",
+    lower: int,
+    upper: int,
+    center: Optional[int] = None,
+    forced: Optional[int] = None,
 ) -> int:
     assert lower <= upper
+    assert forced is None or lower <= forced <= upper
     if lower == upper:
         # Write a value even when this is trivial so that when a bound depends
         # on other values we don't suddenly disappear when the gap shrinks to
@@ -83,7 +88,8 @@ def integer_range(
     elif center == lower:
         above = True
     else:
-        above = not boolean(data)
+        force_above = None if forced is None else forced < center
+        above = not data.draw_bits(1, forced=force_above)
 
     if above:
         gap = upper - center
@@ -95,7 +101,7 @@ def integer_range(
     bits = gap.bit_length()
     probe = gap + 1
 
-    if bits > 24 and data.draw_bits(3):
+    if bits > 24 and data.draw_bits(3, forced=None if forced is None else 0):
         # For large ranges, we combine the uniform random distribution from draw_bits
         # with a weighting scheme with moderate chance.  Cutoff at 2 ** 24 so that our
         # choice of unicode characters is uniform but the 32bit distribution is not.
@@ -104,7 +110,9 @@ def integer_range(
 
     while probe > gap:
         data.start_example(INTEGER_RANGE_DRAW_LABEL)
-        probe = data.draw_bits(bits)
+        probe = data.draw_bits(
+            bits, forced=None if forced is None else abs(forced - center)
+        )
         data.stop_example(discard=probe > gap)
 
     if above:
@@ -113,7 +121,8 @@ def integer_range(
         result = center - probe
 
     assert lower <= result <= upper
-    return int(result)
+    assert forced is None or result == forced, (result, forced, center, above)
+    return result
 
 
 T = TypeVar("T")
@@ -161,10 +170,6 @@ def fractional_float(data: "ConjectureData") -> float:
     return (int_to_float(FLOAT_PREFIX | data.draw_bits(52)) - 1) / FULL_FLOAT
 
 
-def boolean(data: "ConjectureData") -> bool:
-    return bool(data.draw_bits(1))
-
-
 def biased_coin(
     data: "ConjectureData", p: float, *, forced: Optional[bool] = None
 ) -> bool:

hypothesis-python/src/hypothesis/strategies/_internal/numbers.py

@@ -94,7 +94,19 @@ def do_draw(self, data):
                     data.stop_example(discard=probe < self.start)
                 return probe
 
-        return d.integer_range(data, self.start, self.end, center=0)
+        # For bounded integers, make the bounds and near-bounds more likely.
+        forced = None
+        if self.end - self.start > 127:
+            forced = {
+                122: self.start,
+                123: self.start,
+                124: self.end,
+                125: self.end,
+                126: self.start + 1,
+                127: self.end - 1,
+            }.get(data.draw_bits(7))
+
+        return d.integer_range(data, self.start, self.end, center=0, forced=forced)
 
     def filter(self, condition):
         if condition is math.isfinite:

hypothesis-python/tests/array_api/test_arrays.py

@@ -444,8 +444,7 @@ def test_excluded_min_in_float_arrays(xp, xps, dtype, low, data):
 @st.composite
 def distinct_int64_integers(draw):
     used = draw(st.shared(st.builds(set), key="distinct_int64_integers.used"))
-    int64_max = 2**63
-    i = draw(st.integers(1 - int64_max, int64_max).filter(lambda x: x not in used))
+    i = draw(st.integers(-(2**63), 2**63 - 1).filter(lambda x: x not in used))
     used.add(i)
     return i
 

hypothesis-python/tests/cover/test_filtered_strategy.py

@@ -17,7 +17,7 @@ def test_filter_iterations_are_marked_as_discarded():
     variable_equal_to_zero = 0  # non-local references disables filter-rewriting
     x = st.integers(0, 255).filter(lambda x: x == variable_equal_to_zero)
 
-    data = ConjectureData.for_buffer([2, 1, 0])
+    data = ConjectureData.for_buffer([0, 2, 1, 0])
 
     assert data.draw(x) == 0
 

hypothesis-python/tests/quality/test_integers.py

@@ -111,3 +111,20 @@ def f(data):
     bytes_needed = actual_bits_needed // 8
     # 3 extra bytes: two for the sampler, one for the capping value.
     assert len(v.buffer) == 3 + bytes_needed
+
+
+def test_generates_boundary_values_even_when_unlikely():
+    r = Random()
+    trillion = 10**12
+    strat = st.integers(-trillion, trillion)
+    boundary_vals = {-trillion, -trillion + 1, trillion - 1, trillion}
+    for _ in range(10_000):
+        buffer = bytes(r.randrange(0, 255) for _ in range(1000))
+        val = ConjectureData.for_buffer(buffer).draw(strat)
+        boundary_vals.discard(val)
+        if not boundary_vals:
+            break
+    else:
+        raise AssertionError(
+            f"Expected to see all boundary vals, but still have {boundary_vals}"
+        )