Support unindexed VCFs #337
Conversation
|
Oh, that's neat. Thanks @jeromekelleher. From my reading here this is supporting non-indexed VCFs, rather than simply non-compressed VCFs, right? |
|
Compressed or uncompressed is fine. |
|
You're right though, it's the indexing that matters |
jeromekelleher
force-pushed
the
uncompressed-vcf
branch
from
c2b7a63 to
123a9c7
Compare
|
OK, I think this is ready to go. We can now support VCFs that don't have an index and (therefore) can also support uncompressed VCF text. The rules are
Other than that, everything works as before. We can mixed indexed and unindexed files no problem. |
|
Woohoo. This is great. @AprilWei001 will be pleased. |
bio2zarr/vcf_utils.py
Outdated
|
|
||
| def contig_record_counts(self): | ||
| if self.index is None: | ||
| return {self.sequence_names[0]: np.inf} |
There was a problem hiding this comment.
Excellent question! Fixed to use the correct constant.
bio2zarr/vcf_utils.py
Outdated
| TABIX = ".tbi" | ||
|
|
||
|
|
||
| class IndexedVcf(contextlib.AbstractContextManager): |
There was a problem hiding this comment.
The name is now a bit misleading since the VCF is not necessarily indexed. The class is used for partitioning, iterating over variants, and getting metadata about number of variants. I can't think of a better name though!
There was a problem hiding this comment.
Yes, good point. I've renamed to "VcfFile". It's good enough.
Supporting uncompressed VCF is a bit of a faff, but worth the trouble. Here's a WIP for that.
cc @hyanwong