importpkg: support ssdeep hash

2 files changed, 22 insertions, 3 deletions

diff --git a/README b/README
index bc1b715..4e2083c 100644
--- a/README
+++ b/README
@@ -3,6 +3,8 @@ Required packages
 
     aptitude install python python-debian python-lzma curl python-jinja2 python-werkzeug sqlite3 python-imaging
 
+    Optional: https://pypi.python.org/pypi/ssdeep
+
 Create a database
 -----------------
 The database name is currently hardcoded as `test.sqlite3`. So copy the SQL
diff --git a/importpkg.py b/importpkg.py
index d63b85e..84cdad1 100755
--- a/importpkg.py
+++ b/importpkg.py
@@ -11,7 +11,12 @@ from debian.debian_support import version_compare
 from debian import deb822
 import lzma
 
-from dedup.hashing import HashBlacklist, DecompressedHash, SuppressingHash, hash_file
+try:
+    import ssdeep
+except ImportError:
+    ssdeep = None
+
+from dedup.hashing import HashBlacklist, DecompressedHash, SuppressingHash, StoredHash, hash_file
 from dedup.compression import GzipDecompressor, DecompressedStream
 from dedup.image import ImageHash
 
@@ -96,12 +101,24 @@ def imagehash():
     hashobj.name = "image_sha512"
     return hashobj
 
+if ssdeep is None:
+    def multihash():
+        return MultiHash(sha512_nontrivial(), gziphash(), imagehash())
+else:
+    def ssdeephash():
+        hashobj = StoredHash(lambda bytesio: ssdeep.hash(bytesio.getvalue()))
+        hashobj.name = "ssdeep"
+        return hashobj
+
+    def multihash():
+        return MultiHash(sha512_nontrivial(), gziphash(), imagehash(),
+                         ssdeephash())
+
 def get_hashes(tar):
     for elem in tar:
         if not elem.isreg(): # excludes hard links as well
             continue
-        hasher = MultiHash(sha512_nontrivial(), gziphash(), imagehash())
-        hasher = hash_file(hasher, tar.extractfile(elem))
+        hasher = hash_file(multihash(), tar.extractfile(elem))
         for hashobj in hasher.hashes:
             hashvalue = hashobj.hexdigest()
             if hashvalue: