Fix overflow in searchsorted*
#286
Conversation
Codecov Report
@@ Coverage Diff @@
## master #286 +/- ##
==========================================
- Coverage 95.90% 95.53% -0.37%
==========================================
Files 5 5
Lines 464 426 -38
==========================================
- Hits 445 407 -38
Misses 19 19
Continue to review full report at Codecov.
|
|
Hmmm, the failure on 1.0 is concerning. Do you know why that happens? |
|
Yes, should be fixed now. I had made a mistake in the test, but since empty ranges are equal in 1.6+, the test had passed. |
|
I feel like we're unnecessarily complicating matters here. We should simply pass the |
|
I think the codecov changes are ok, looks good to merge |
jishnub
changed the title
src/OffsetArrays.jl
Outdated
| for f in [:searchsortedfirst, :searchsortedlast, :searchsorted] | ||
| _safe_f = Symbol("_safe_" * String(f)) | ||
| @eval function $_safe_f(v::OffsetArray, x, ilo, ihi, o::Base.Ordering) | ||
| offset = firstindex(v) - firstindex(parent(v)) |
There was a problem hiding this comment.
What if v is an offset vector of another offset vector? E.g., OffsetVector of PaddedView?
Edit: oh never mind, it seems to work just fine by delegating to the parent method.
There was a problem hiding this comment.
Maybe I'll add some tests, good to be sure
Fixes #285, but this doesn't fix all issues (as certain tests are broken). However, I think
searchsortedfirstandsearchsortedlastshould be fixed.