Closed hatf0 closed 2 years ago
Merging #4 (87c352e) into master (a1a0abe) will increase coverage by
0.56%
. The diff coverage is99.15%
.
@@ Coverage Diff @@
## master #4 +/- ##
==========================================
+ Coverage 84.88% 85.45% +0.56%
==========================================
Files 31 32 +1
Lines 6624 6902 +278
==========================================
+ Hits 5623 5898 +275
- Misses 1001 1004 +3
Impacted Files | Coverage Δ | |
---|---|---|
source/mir/ion/ser/msgpack.d | 98.91% <99.15%> (ø) |
|
source/mir/ion/deser/json.d | 84.21% <0.00%> (ø) |
|
source/mir/ion/ser/package.d | 96.72% <0.00%> (+0.01%) |
:arrow_up: |
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 a1a0abe...87c352e. Read the comment docs.
Do not merge yet -- I have to refactor some bits and pull out the usage of ScopedBuffer (as it's now going out of scope and causing a use after free)
Please add extra 4 spaces for the whole MsgpackSerializer. And use the same order of methods like in the original code. That is needed to reduce the diff for easier review.
We need 100% coverage of the new code for the message pack. It is a critical part of code used in production.
We need 100% coverage of the new code for the message pack. It is a critical part of code used in production.
@9il, I was able to get coverage up to 98% on this file, if that's fine. The only code that isn't covered is:
https://github.com/libmir/mir-ion/blob/f143764fc10701beda287f72cc38eec0042f5b8a/source/mir/ion/ser/msgpack.d#L572-L575
https://github.com/libmir/mir-ion/blob/f143764fc10701beda287f72cc38eec0042f5b8a/source/mir/ion/ser/msgpack.d#L537-L540
Both of these aren't feasible to test (as they'd require that we allocate a const(char)[]
array that is 4.29 GB in size)
https://github.com/libmir/mir-ion/blob/f143764fc10701beda287f72cc38eec0042f5b8a/source/mir/ion/ser/msgpack.d#L317-L318
and I'm not sure how to trigger this code path.
@9il, I was able to get coverage up to 98% on this file, if that's fine. The only code that isn't covered is:
That is enough, thank you
Alright. I've added your requested fixes -- let me know if there's anything else that need fixing.
Looks good.
This PR will fully remove the dependency on
msgpack-d
for the serialization backend. All features should be implemented (and are verified to be correct with the upstream MessagePack spec).Let me know if this needs any more work! Thanks.