Add BGZFCodec for reading and writing files with a .bgz suffix. #106
Conversation
|
I'm generally in favor of this, though it may take us a while to test. |
|
I'd like this to go into the next release, so please take a look/try it out @heuermh. |
| private static VCFFileReader parseVcf(File vcf) throws IOException { | ||
| File actualVcf; | ||
| // work around TribbleIndexedFeatureReader not reading header from .bgz files | ||
| if (vcf.getName().endsWith(".bgz")) { |
There was a problem hiding this comment.
Is this only a workaround for unit tests, or something that we'll also need to do downstream?
There was a problem hiding this comment.
This is only needed for testing. VCFHeaderReader in Hadoop-BAM will correctly read a bgzf stream now (since #97).
There was a problem hiding this comment.
Great, thanks! Yep, VCFHeaderReader is now working for us.
There was a problem hiding this comment.
@tomwhite - is that (.bgz/TribbleIndexedFeatureReader) a known issue or is there a ticket? It seems like something we should fix for the future....
There was a problem hiding this comment.
I just created samtools/htsjdk#653 for it.
|
+1, see downstream pull request referenced above |
| String extension = ""; | ||
| if (isCompressed) { | ||
| Class<? extends CompressionCodec> codecClass = | ||
| getOutputCompressorClass(ctx, GzipCodec.class); |
There was a problem hiding this comment.
GzipCodec seems like an unnatural default, certainly for BCF. Wouldn't BGZFCodec be a better choice for both for both VCF and BCF?
|
A couple of minor comments. Not sure why the Travis push job failed. Otherwise LGTM. Back to @tomwhite. |
The old BGZF codec has been renamed BGZFEnhancedGzipCodec and reads .gz files as BGZF (so they are splittable) if possible, otherwise falls back to regular gzip.
|
Thank you, @tomwhite! |
a .gz suffix), see HadoopGenomics/Hadoop-BAM#106. Also remove spurious debug.
a .gz suffix), see HadoopGenomics/Hadoop-BAM#106. Also remove spurious debug.
a .gz suffix), see HadoopGenomics/Hadoop-BAM#106. Also remove spurious debug.
This adds support for writing BGZF-compressed files.
The old BGZF codec has been renamed BGZFEnhancedGzipCodec and reads .gz files
as BGZF (so they are splittable) if possible, otherwise falls back to regular
gzip.