Closed charlesgregory closed 3 years ago
Merging #88 (2955703) into develop (3f34c1b) will decrease coverage by
0.62%
. The diff coverage is90.66%
.
@@ Coverage Diff @@
## develop #88 +/- ##
===========================================
- Coverage 80.53% 79.90% -0.63%
===========================================
Files 40 40
Lines 2805 2832 +27
===========================================
+ Hits 2259 2263 +4
- Misses 546 569 +23
Impacted Files | Coverage Δ | |
---|---|---|
source/dhtslib/bed/reader.d | 100.00% <ø> (ø) |
|
source/dhtslib/bed/record.d | 63.04% <ø> (ø) |
|
source/dhtslib/gff/record.d | 82.25% <ø> (ø) |
|
source/dhtslib/sam/record.d | 81.25% <0.00%> (ø) |
|
source/dhtslib/vcf/record.d | 91.21% <50.00%> (ø) |
|
source/dhtslib/coordinates.d | 97.29% <93.61%> (-1.67%) |
:arrow_down: |
source/dhtslib/bed/package.d | 100.00% <100.00%> (ø) |
|
source/dhtslib/faidx.d | 89.47% <100.00%> (+0.58%) |
:arrow_up: |
source/dhtslib/gff/reader.d | 94.28% <100.00%> (+0.16%) |
:arrow_up: |
source/dhtslib/sam/reader.d | 86.00% <100.00%> (-0.23%) |
:arrow_down: |
... and 6 more |
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 3f34c1b...2955703. Read the comment docs.
For ChromCoordinates could you use htslib region parsing code for a potential speed boost (and bug freeness?) Or too hard to use?
Potentially, I will look into htslib and see if that function is exported and how it could be used. Probably more thoroughly tested so its worth a look. Honestly, I need to overload the region structs in dhtslib.sam.reader
we did a while ago.
In addition to prior comment on htslib's string parsing which i'd like you to answer, I am also wondering about the use of Coordinate and Coordinates which seems confusing inasmuch as their names are almost identical. It may be too late int he game to change the latter, but what about Interval or Range instead?
The nice thing about version control is any prior implementations we are already using will still work. I am open to changes as long as we don't change the API too often. I would rather we have the best implementation before the paper is out as more people start looking at the code afterwards. After we are done, we can just bump the version to 0.13 or even move to version 1.0.
For ChromCoordinates could you use htslib region parsing code for a potential speed boost (and bug freeness?) Or too hard to use?
Potentially, I will look into htslib and see if that function is exported and how it could be used. Probably more thoroughly tested so its worth a look. Honestly, I need to overload the region structs in
dhtslib.sam.reader
we did a while ago.
Great news:
Latest commit changes Coordinates
to Interval
and removes ChromCoordinates
.
@charlesgregory ChromInterval
still exists in current version of PR as type parameter to some function calls in e.g. sam/
; however, was this type removed ?
I have addressed the remaining concerns in the last two commits. Since we are likely to change this implementation further anyway, I am going to accept this pull request so I can begin further work.
Fixes #81, #82, #85.
ZBHO val = ZBHO(1, 0);
ChromCoordinates
shortcuts are of form:ChromZBHO
Coordinates
.OBC newVal = val.to!OBC;