[bugfix] Fix physical key pruning issue

Fixes hail-is#4874
tpoterba · Apr 16, 2019 · fbf7b28 · fbf7b28
1 parent 3314180
commit fbf7b28
Show file tree

Hide file tree

Showing 2 changed files with 8 additions and 2 deletions.
diff --git a/hail/python/test/hail/table/test_table.py b/hail/python/test/hail/table/test_table.py
@@ -1048,6 +1048,11 @@ def test_unicode_ordering(self):
         ht = ht.annotate(fd=hl.sorted(a))
         assert ht.fd.collect()[0] == ["e", "é"]
 
+    def test_physical_key_truncation(self):
+        path = new_temp_file(suffix='ht')
+        hl.import_vcf(resource('sample.vcf')).rows().key_by('locus').write(path)
+        hl.read_table(path).select()._force_count()
+
 def test_large_number_of_fields(tmpdir):
     ht = hl.utils.range_table(100)
     ht = ht.annotate(**{

diff --git a/hail/src/main/scala/is/hail/rvd/AbstractRVDSpec.scala b/hail/src/main/scala/is/hail/rvd/AbstractRVDSpec.scala
@@ -139,9 +139,10 @@ abstract class AbstractRVDSpec {
   def codecSpec: CodecSpec
 
   def read(hc: HailContext, path: String, requestedType: PStruct): RVD = {
-    val rvdType = RVDType(requestedType, key)
+    val requestedKey = key.takeWhile(requestedType.hasField)
+    val rvdType = RVDType(requestedType, requestedKey)
 
-    RVD(rvdType, partitioner, hc.readRows(path, encodedType, codecSpec, partFiles, requestedType))
+    RVD(rvdType, partitioner.coarsen(requestedKey.length), hc.readRows(path, encodedType, codecSpec, partFiles, requestedType))
   }
 
   def readLocal(hc: HailContext, path: String, requestedType: PStruct): IndexedSeq[Row] =