Add workaround for BAM files hosted on SRA s3 archive.

bigdatagenomics · May 10, 2022 · 4614261 · 4614261
1 parent 4fa2b08
commit 4614261
Show file tree

Hide file tree

Showing 2 changed files with 21 additions and 3 deletions.
diff --git a/adam-core/src/main/scala/org/bdgenomics/adam/util/FileExtensions.scala b/adam-core/src/main/scala/org/bdgenomics/adam/util/FileExtensions.scala
@@ -35,9 +35,11 @@ object FileExtensions {
    * @return Returns true if the path name matches a BAM/CRAM/SAM format file extension.
    */
   def isBamExt(pathName: String): Boolean = {
-    pathName.endsWith(".bam") ||
-      pathName.endsWith(".cram") ||
-      pathName.endsWith(".sam")
+    // workaround for BAM files hosted on SRA s3 archive
+    val trimmed = pathName.replaceAll("\\.[0-9]+$", "")
+    trimmed.endsWith(".bam") ||
+      trimmed.endsWith(".cram") ||
+      trimmed.endsWith(".sam")
   }
 
   /**

diff --git a/adam-core/src/test/scala/org/bdgenomics/adam/util/FileExtensionsSuite.scala b/adam-core/src/test/scala/org/bdgenomics/adam/util/FileExtensionsSuite.scala
@@ -36,4 +36,20 @@ class FileExtensionsSuite extends FunSuite {
     assert(FileExtensions.isVcfExt("file.vcf.gz"))
     assert(!FileExtensions.isVcfExt("file.vcf.bgzf"))
   }
+
+  test("is a bam/cram/sam extension") {
+    assert(FileExtensions.isBamExt("file.bam"))
+    assert(FileExtensions.isBamExt("file.cram"))
+    assert(FileExtensions.isBamExt("file.cram"))
+    assert(!FileExtensions.isBamExt("file.vcf"))
+    assert(FileExtensions.isBamExt("file.bam.1"))
+    assert(FileExtensions.isBamExt("file.bam.10"))
+    assert(FileExtensions.isBamExt("file.cram.1"))
+    assert(FileExtensions.isBamExt("file.cram.10"))
+    assert(FileExtensions.isBamExt("file.sam.1"))
+    assert(FileExtensions.isBamExt("file.sam.10"))
+    assert(!FileExtensions.isBamExt("file.bam.x"))
+    assert(!FileExtensions.isBamExt("file.cram.y"))
+    assert(!FileExtensions.isBamExt("file.sam.z"))
+  }
 }