GitLab

core/segment.cpp: KmerMultiSegmenter keeps the 'best unsegmented' kseg when the sequence is not segmented

This allows to report a more meaningful unsegmentation cause than just picking the last one.
Let say that the first point of the current CHANGELOG already included this feature.

We now store information either on lengths (on reads) or on number of reads (on clones).

We store in the 'clone_stats' member of germlines the statistics related to clones
(number of clones, number of reads per clone).

.json elements are sorted now by keys.

Completes 4045ed90

With the updated multi-germline heurisitic (ad552cb7), selecting -i is safer now.
See in particular tests/multi-i(-short).should_get (2ade70be).

Stuttering?

The format has evolved. Test updated accordingly

format_json.py raises an exception if the JSON is not valid.
Using format_json.py needs to slightly rewrite the test
(spaces are deleted, and the order may not be the same).

If the JSON is not valid, raises an exception

format-json.sh is deleted because format_json.py has
similar features while checking syntax additionnally.

    Tests were failed because fuse.py was not packaged with the release.
    This is not the case anymore. So the tests can be un-failed.

ReadChooser doesn't use a file.

ReadChooser was storing all the sequences and score in a map (log-time access)
and then reads were stored in a vector which was sorted depending on the
scores stored in the map. So we had many allocations (map+vector) and a slow
access (log access time, so the sort was in n log²n, where n is the number of
sequences). I'm a bit ashamed to have done such an inefficient thing.

Now the scores and the sequences are stored in an array. The sequences
are not copied, but just a pointer to them is stored.

The C function qsort is used because the C++ sort doesn't work
(because of implicit destruction of something).

The code is now much quicker (a -c clones of 600k reads, with -z 1 is now twice as fast).

Someone motivated could probably do better with a radix sort.

For the moment the representative doesn't compute
a quality.

Since we copy the information from the best sequence if the sequence had
quality information, we were also getting it for the representative... and it
was displayed in the .vdj.fa file.

C++11-ready.

Append some flags to CXXFLAGS
and use LDLIBS instead of LDFLAGS for -lm and -lz

See https://www.gnu.org/software/make/manual/html_node/Implicit-Variables.html

Affine gaps are enabled only when using some Costs such as VDJaffine.

Implements affine gaps by storing three matrices, B, Bins and Bdel.