Closed mejrpete closed 3 years ago
Merging #97 into master will decrease coverage by
0.50%
. The diff coverage isn/a
.
@@ Coverage Diff @@
## master #97 +/- ##
==========================================
- Coverage 98.30% 97.80% -0.51%
==========================================
Files 47 48 +1
Lines 3014 3054 +40
Branches 114 121 +7
==========================================
+ Hits 2963 2987 +24
- Misses 41 51 +10
- Partials 10 16 +6
Impacted Files | Coverage Δ | |
---|---|---|
tests/io/common.py | 60.00% <0.00%> (ø) |
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 2d57d46...8f9e85c. Read the comment docs.
Hi @mejrpete , thanks for the PR. We are definitely interested in supporting macos, though I have to say that at the moment I don't have access to any such machine :-)
But this looks definitely good. For the record, it might be good to add osx testing to our travis tests, e.g. as described in https://docs.travis-ci.com/user/multi-os/.
I'm not sure how interested you all are in supporting
tarski
on macos. I know that a few people working with it now do development work with it locally on macs, although I believe it's generally deployed in linux environments.Previously, tests (as well as actual use) would fail on macos due to a mismatch between the expected cytpe ID for
stdout
(stdout
-- as it typically is on linux) and the actual ctype ID (__stdoutp
). This fix just directly assigns the string to be__stdoutp
in the case when we are in adarwin
environment.As far as I can tell, this is the best way to do this that's straightforward, since it appears that
stdout
is generally a macro, rather than a variable. It appears macos is the weird one out here, so it seems to make sense to just fix the problem directly.