Merge pull request #13523 from velconia/fix_rnn_search

Fix reader of rnn_search in python3

Author: velconia

python/paddle/dataset/wmt14.py

@@ -89,7 +89,8 @@ def reader():
             ]
             for name in names:
                 for line in f.extractfile(name):
-                    line_split = line.strip().split(six.b('\t'))
+                    line = cpt.to_text(line)
+                    line_split = line.strip().split('\t')
                     if len(line_split) != 2:
                         continue
                     src_seq = line_split[0]  # one source sequence

python/paddle/dataset/wmt16.py

@@ -64,7 +64,8 @@ def __build_dict(tar_file, dict_size, save_path, lang):
     word_dict = defaultdict(int)
     with tarfile.open(tar_file, mode="r") as f:
         for line in f.extractfile("wmt16/train"):
-            line_split = line.strip().split(six.b("\t"))
+            line = cpt.to_text(line)
+            line_split = line.strip().split("\t")
             if len(line_split) != 2: continue
             sen = line_split[0] if lang == "en" else line_split[1]
             for w in sen.split():
@@ -123,7 +124,8 @@ def reader():
 
         with tarfile.open(tar_file, mode="r") as f:
             for line in f.extractfile(file_name):
-                line_split = line.strip().split(six.b("\t"))
+                line = cpt.to_text(line)
+                line_split = line.strip().split("\t")
                 if len(line_split) != 2:
                     continue
                 src_words = line_split[src_col].split()