diff options
| -rw-r--r-- | python/refcat/tasks.py | 23 |
1 files changed, 21 insertions, 2 deletions
diff --git a/python/refcat/tasks.py b/python/refcat/tasks.py index f131fbc..7c1fabc 100644 --- a/python/refcat/tasks.py +++ b/python/refcat/tasks.py @@ -863,18 +863,37 @@ class UnmatchedMapped(Refcat): return luigi.LocalTarget(path=self.path(ext="tsv.zst"), format=Zstd) +class UnmatchedRefsToRelease(Refcat): + """ + Convert unmatched refs to releases. + """ + def requires(self): + return UnmatchedRefs() + + def run(self): + output = shellout(""" + zstdcat -T0 {input} | + skate-conv -f ref | + zstd -T0 -c > {output} + """, + 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 UnmatchedResolveJournalNames(Refcat): """ Try to resolve journal names so we can match against both abbreviations. Keep only the resolved docs (for now). """ def requires(self): - return UnmatchedRefs() + return UnmatchedRefsToRelease() def run(self): output = shellout(""" zstdcat -T0 {input} | - skate-conv -f ref | skate-resolve-journal-name -R -f 1 -B -A {abbrev} | zstd -T0 -c > {output} """, |