Make SignedShrinker bounded for <$ty>::MIN #296
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Previously, the `SignedShrinker` was guaranteed to yield a result if initialized with `<$ty>::MIN` (e.g. `i32::MIN` for shrinking `i32` and `f32`). This would result in endless shrinking if a test would happen to fail for these values. Shrinking signed values is done by halfing a second value `i` and subtracting it from the original `x`. Thus `i` will be equal to `0` at some point. We choose to halt the iteration in this state, i.e. not yield any new values. Yielding `x - 0` would be pointless anyway, since that obviously equals the original value which was the initial witness for the test failure.
|
Maybe it makes sense to introduce shrinking tests which assert that the shrinkers included in this crate yield at most $number values? |
The condition `(self.x - self.i).abs() < self.x.abs()` is true if `self.i` is not `0` and has the same sign than `self.x`. The latter is always the case due to the initialization, the former was recently introduced to bound the iteration in the case of `self.x == <$ty>::MIN`. Thus, both of these conditions became redundant.
For floating point values, we use a shrinker for signed integers and convert the output to the float's type. The underlying shrinker is bounded and never includes the original value in the output, i.e. it behaves correctly. However, when converting the value to the target type we may loose some information and thus yield the original value, again. If the test happens to fail for such a value, we may end up endlessly repeating the test with a sequence of values including the original one over and over again. This change avoids the issue by applying an additional filter.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Previously, the
SignedShrinkerwas guaranteed to yield a result if initialized with<$ty>::MIN(e.g.i32::MINfor shrinkingi32andf32). This would result in endless shrinking if a test would happen to fail for these values.Shrinking signed values is done by halfing a second value
iand subtracting it from the originalx. Thusiwill be equal to0at some point. We choose to halt the iteration in this state, i.e. not yield any new values. Yieldingx - 0would be pointless anyway, since that obviously equals the original value which was the initial witness for the test failure.Closes #295, #301.