tasks: add exact join tasks

1 files changed, 48 insertions, 2 deletions

diff --git a/python/refcat/tasks.py b/python/refcat/tasks.py
index 9d4c982..07c1217 100644
--- a/python/refcat/tasks.py
+++ b/python/refcat/tasks.py
@@ -572,8 +572,8 @@ class BrefZipPMID(Refcat):
     """
     def requires(self):
         return {
-            "refs": RefsDOI(),
-            "fatcat": FatcatDOI(),
+            "refs": RefsPMID(),
+            "fatcat": FatcatPMID(),
         }
 
     def run(self):
@@ -587,3 +587,49 @@ class BrefZipPMID(Refcat):
 
     def output(self):
         return luigi.LocalTarget(path=self.path(ext="json.zst"), format=Zstd)
+
+
+class BrefZipPMCID(Refcat):
+    """
+    Run skate-reduce from two files.
+    """
+    def requires(self):
+        return {
+            "refs": RefsPMCID(),
+            "fatcat": FatcatPMCID(),
+        }
+
+    def run(self):
+        output = shellout(r"""
+                          skate-reduce -m exact -r pmcid -F <(zstdcat -T0 {refs}) -L <(zstdcat -T0 {fatcat}) |
+                          zstd -c -T0 > {output}
+                          """,
+                          refs=self.input().get("refs").path,
+                          fatcat=self.input().get("fatcat").path)
+        luigi.LocalTarget(output).move(self.output().path)
+
+    def output(self):
+        return luigi.LocalTarget(path=self.path(ext="json.zst"), format=Zstd)
+
+
+class BrefZipArxiv(Refcat):
+    """
+    Run skate-reduce from two files.
+    """
+    def requires(self):
+        return {
+            "refs": RefsArxiv(),
+            "fatcat": FatcatArxiv(),
+        }
+
+    def run(self):
+        output = shellout(r"""
+                          skate-reduce -m exact -r arxiv -F <(zstdcat -T0 {refs}) -L <(zstdcat -T0 {fatcat}) |
+                          zstd -c -T0 > {output}
+                          """,
+                          refs=self.input().get("refs").path,
+                          fatcat=self.input().get("fatcat").path)
+        luigi.LocalTarget(output).move(self.output().path)
+
+    def output(self):
+        return luigi.LocalTarget(path=self.path(ext="json.zst"), format=Zstd)