Closed tlienart closed 4 years ago
Great, hasproperty
is not defined for Julia 1... come on...
Merging #55 into master will decrease coverage by
3.4%
. The diff coverage is96.66%
.
@@ Coverage Diff @@
## master #55 +/- ##
==========================================
- Coverage 93.56% 90.16% -3.41%
==========================================
Files 6 6
Lines 171 183 +12
==========================================
+ Hits 160 165 +5
- Misses 11 18 +7
Impacted Files | Coverage Δ | |
---|---|---|
src/ScientificTypes.jl | 85.71% <ø> (ø) |
:arrow_up: |
src/conventions/mlj/finite.jl | 83.78% <100%> (-16.22%) |
:arrow_down: |
src/conventions/mlj/mlj.jl | 93.1% <100%> (+0.79%) |
:arrow_up: |
src/autotype.jl | 97.87% <100%> (ø) |
:arrow_up: |
src/tables.jl | 88.46% <66.66%> (-2.85%) |
: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 fe7e105...22bc21e. Read the comment docs.
Agreed!!
"hasproperty"-dispatch appropriate in this case
This is a hack that solves the issue discussed in #49, it could be temporarily implemented while we wait to get a better way from https://github.com/JuliaData/Tables.jl/issues/122
The hack is that I don't want to add LazyArrays as a dependency, so I can't explicitly check that things are LazyArrays, but I can check their properties, and an
ApplyArray
has 2 properties::f
and:args
, the bet is that temporarily this is enough to uniquely identify such an array, in which case a conversion to a hard Vector is made.This solves the issue in #49 on my side and I would suggest we bring that in while waiting for a better solution? (unless there is already a better solution)