Closed rafaqz closed 3 years ago
Merging #43 (8b01aed) into main (5385cc9) will decrease coverage by
0.07%
. The diff coverage is81.81%
.
@@ Coverage Diff @@
## main #43 +/- ##
==========================================
- Coverage 85.95% 85.87% -0.08%
==========================================
Files 10 9 -1
Lines 178 177 -1
==========================================
- Hits 153 152 -1
Misses 25 25
Flag | Coverage Δ | |
---|---|---|
unittests | 85.87% <81.81%> (-0.08%) |
:arrow_down: |
Flags with carried forward coverage won't be shown. Click here to find out more.
Impacted Files | Coverage Δ | |
---|---|---|
src/tables.jl | 66.66% <66.66%> (ø) |
|
src/types/GBIFRecords.jl | 90.62% <87.50%> (-1.05%) |
:arrow_down: |
Continue to review full report at Codecov.
Legend - Click here to learn more
Δ = absolute <relative> (impact)
,ø = not affected
,? = missing data
Powered by Codecov. Last update 5385cc9...8b01aed. Read the comment docs.
Let's re-open this when the Tables interface is fixed, as the PR currently mixes the two. There's a reason why this line is here, I'll dig it up in the meantime.
Maybe just refactor out the Tables part and reopen? The type piracy part seems important
I must have rebased that tables code in by accident. Let me know next time before closing, as it wasn't possible to reopen this after a force-pushing the fix to this branch.
I made a new PR in #48
Adding
convert(::AbstractString, ::Nothing)
is type piracy on core julia types. It affects every package loaded in a session and all of Base Julia.. The main issue being that it will allow code to work on a system with GBIF.jl loaded that will fail on another system without GBIF loaded, which is not the worst kind of piracy, but we still can't.We should really find where
nothing
is being returned, and wrap it with check that inserts missing instead. Is there an example taxon that triggers the bug?