decouple a function decompress out of decompress_tar

Building on the previous commit, add a decompress function that turns a
compressed filelike into a decompressed filelike. Use it to decouple the
decompression step.

1 files changed, 3 insertions, 10 deletions

diff --git a/importpkg.py b/importpkg.py
index 01ec87a..dac4bb1 100755
--- a/importpkg.py
+++ b/importpkg.py
@@ -11,13 +11,12 @@ import sys
 import tarfile
 import zlib
 
-import lzma
 import yaml
 
 from dedup.debpkg import DebExtractor, process_control, get_tar_hashes
 from dedup.hashing import DecompressedHash, SuppressingHash, HashedStream, \
         HashBlacklistContent
-from dedup.compression import GzipDecompressor, DecompressedStream
+from dedup.compression import GzipDecompressor, decompress
 from dedup.image import GIFHash, PNGHash
 
 boring_content = set(("", "\n"))
@@ -44,14 +43,8 @@ def gifhash():
     return hashobj
 
 def decompress_tar(filelike, extension):
-    if extension in (b".lzma", b".xz"):
-        filelike = DecompressedStream(filelike, lzma.LZMADecompressor())
-        extension = b""
-    if extension not in (b"", b".gz", b".bz2"):
-        raise ValueError("unknown compression format with extension %r" %
-                         extension)
-    return tarfile.open(fileobj=filelike,
-                        mode="r|" + extension[1:].decode("ascii"))
+    filelike = decompress(filelike, extension.decode("ascii"))
+    return tarfile.open(fileobj=filelike, mode="r|")
 
 class ProcessingFinished(Exception):
     pass