From 9cdef6270986b48746fa7e4a610f1bbeb2ed8d3d Mon Sep 17 00:00:00 2001 From: Martin Czygan Date: Wed, 19 May 2021 22:33:01 +0200 Subject: tasks: add RefsMapped --- python/refcat/tasks.py | 26 ++++++++++++++++++++++++++ 1 file changed, 26 insertions(+) (limited to 'python') diff --git a/python/refcat/tasks.py b/python/refcat/tasks.py index b00302e..6d7bf56 100644 --- a/python/refcat/tasks.py +++ b/python/refcat/tasks.py @@ -543,6 +543,32 @@ class FatcatMapped(Refcat): return luigi.LocalTarget(path=self.path(ext="tsv.zst"), format=Zstd) +class RefsMapped(Refcat): + """ + Apply mapper on refs. + """ + mapper = luigi.Parameter(default="ts", description="mapper short name") + + def requires(self): + return RefsToRelease() + + def run(self): + output = shellout(""" + zstdcat -T0 {input} | + skate-map -m {mapper} -skip-on-empty 1 | + LC_ALL=C sort -T {tmpdir} -k1,1 -S25% --parallel 4 | + zstd -T0 -c > {output} + """, + n=self.n, + mapper=self.mapper, + tmpdir=self.tmpdir, + input=self.input().path) + luigi.LocalTarget(output).move(self.output().path) + + def output(self): + return luigi.LocalTarget(path=self.path(ext="tsv.zst"), format=Zstd) + + class BrefZipDOI(Refcat): """ Run skate-reduce from two files. -- cgit v1.2.3