Merge pull request #732 from PaddlePaddle/dataset

fix dataset meta path

examples/dataset/aidatatang_200zh/aidatatang_200zh.py

@@ -102,7 +102,9 @@ def create_manifest(data_dir, manifest_path_prefix):
             for line in json_lines:
                 fout.write(line + '\n')
 
-        with open(dtype + '.meta', 'w') as f:
+        manifest_dir = os.path.dirname(manifest_path_prefix)
+        meta_path = os.path.join(manifest_dir, dtype) + '.meta'
+        with open(meta_path, 'w') as f:
             print(f"{dtype}:", file=f)
             print(f"{total_num} utts", file=f)
             print(f"{total_sec / (60*60)} h", file=f)

examples/dataset/aishell/aishell.py

@@ -103,7 +103,9 @@ def create_manifest(data_dir, manifest_path_prefix):
             for line in json_lines:
                 fout.write(line + '\n')
 
-        with open(dtype + '.meta', 'w') as f:
+        manifest_dir = os.path.dirname(manifest_path_prefix)
+        meta_path = os.path.join(manifest_dir, dtype) + '.meta'
+        with open(meta_path, 'w') as f:
             print(f"{dtype}:", file=f)
             print(f"{total_num} utts", file=f)
             print(f"{total_sec / (60*60)} h", file=f)

examples/dataset/librispeech/librispeech.py

@@ -114,8 +114,10 @@ def create_manifest(data_dir, manifest_path):
         for line in json_lines:
             out_file.write(line + '\n')
 
-    subset = manifest_path
-    with open(subset + '.meta', 'w') as f:
+    subset = os.path.splitext(manifest_path)[1][1:]
+    manifest_dir = os.path.dirname(manifest_path)
+    meta_path = os.path.join(manifest_dir, dtype) + '.meta'
+    with open(meta_path, 'w') as f:
         print(f"{subset}:", file=f)
         print(f"{total_num} utts", file=f)
         print(f"{total_sec / (60*60)} h", file=f)

examples/dataset/mini_librispeech/mini_librispeech.py

@@ -93,8 +93,10 @@ def create_manifest(data_dir, manifest_path):
         for line in json_lines:
             out_file.write(line + '\n')
 
-    subset = os.path.splitext(manifest_path)[1]
-    with open(subset + '.meta', 'w') as f:
+    subset = os.path.splitext(manifest_path)[1][1:]
+    manifest_dir = os.path.dirname(manifest_path)
+    meta_path = os.path.join(manifest_dir, dtype) + '.meta'
+    with open(meta_path, 'w') as f:
         print(f"{subset}:", file=f)
         print(f"{total_num} utts", file=f)
         print(f"{total_sec / (60*60)} h", file=f)

examples/dataset/thchs30/thchs30.py

@@ -139,7 +139,9 @@ def create_manifest(data_dir, manifest_path_prefix):
             for line in json_lines:
                 fout.write(line + '\n')
 
-        with open(dtype + '.meta', 'w') as f:
+        manifest_dir = os.path.dirname(manifest_path_prefix)
+        meta_path = os.path.join(manifest_dir, dtype) + '.meta'
+        with open(meta_path, 'w') as f:
             print(f"{dtype}:", file=f)
             print(f"{total_num} utts", file=f)
             print(f"{total_sec / (60*60)} h", file=f)

examples/dataset/timit/timit.py

@@ -198,7 +198,9 @@ def create_manifest(data_dir, manifest_path_prefix):
             for line in json_lines:
                 fout.write(line + '\n')
 
-        with open(dtype.lower() + '.meta', 'w') as f:
+        manifest_dir = os.path.dirname(manifest_path_prefix)
+        meta_path = os.path.join(manifest_dir, dtype.lower()) + '.meta'
+        with open(meta_path, 'w') as f:
             print(f"{dtype}:", file=f)
             print(f"{total_num} utts", file=f)
             print(f"{total_sec / (60*60)} h", file=f)