Merge pull request #11 from sony/feature/20180416-efficient-create-im…

…agenet-files

Use extractfile in extracting imagenet tar

imagenet-classification/create_train_dir.py

@@ -29,14 +29,11 @@
 dst_dir = args.outdir
 
 with tarfile.open(source_tar_file) as tf:
-    tar_tmp_dir = dst_dir + '/' + 'tmpdir'
-    tf.extractall(tar_tmp_dir)
-
-for tar_file in tqdm.tqdm(os.listdir(tar_tmp_dir)):
-    name, ext = os.path.splitext(os.path.basename(tar_file))
-    category_dir = dst_dir + '/' + name
-    os.mkdir(category_dir)
-    with tarfile.open(tar_tmp_dir + '/' + tar_file) as tf:
-        tf.extractall(category_dir)
-
-shutil.rmtree(tar_tmp_dir)
+    for tar_file_info in tqdm.tqdm(tf.getmembers()):
+        fullname = tar_file_info.name
+        name, ext = os.path.splitext(os.path.basename(fullname))
+        category_dir = dst_dir + '/' + name
+        os.mkdir(category_dir)
+        fileobj = tf.extractfile(tar_file_info)
+        with tarfile.open(fileobj=fileobj) as tf_class:
+            tf_class.extractall(category_dir)