From ef3a19310c3c16c5d956ed19096f908b4502c3e2 Mon Sep 17 00:00:00 2001
From: Steffy-zxf <48793257+Steffy-zxf@users.noreply.github.com>
Date: Fri, 31 May 2019 10:14:29 +0800
Subject: [PATCH] fix-encoding-format (#44)

---
 paddlehub/commands/hub.py |  8 +++++++-
 paddlehub/io/parser.py    | 18 ++++++++++++------
 2 files changed, 19 insertions(+), 7 deletions(-)

diff --git a/paddlehub/commands/hub.py b/paddlehub/commands/hub.py
index c5f901cc..877b3faf 100644
--- a/paddlehub/commands/hub.py
+++ b/paddlehub/commands/hub.py
@@ -58,7 +58,13 @@ command = HubCommand.instance()
 
 
 def main():
-    command.execute(sys.argv[1:])
+    argv = []
+    for item in sys.argv:
+        if six.PY2:
+            argv.append(item.decode(sys.stdin.encoding).decode("utf8"))
+        else:
+            argv.append(item)
+    command.execute(argv[1:])
 
 
 if __name__ == "__main__":
diff --git a/paddlehub/io/parser.py b/paddlehub/io/parser.py
index 4a60105a..a3c7b4aa 100644
--- a/paddlehub/io/parser.py
+++ b/paddlehub/io/parser.py
@@ -17,6 +17,8 @@ from __future__ import absolute_import
 from __future__ import division
 from __future__ import print_function
 
+import codecs
+import sys
 import yaml
 
 
@@ -28,7 +30,7 @@ class CSVFileParser(object):
         pass
 
     def parse(self, csv_file):
-        with open(csv_file, "r") as file:
+        with codecs.open(txt_file, "r", sys.stdin.encoding) as file:
             content = file.read()
         content = content.split('\n')
         self.title = content[0].split(',')
@@ -55,7 +57,7 @@ class YAMLFileParser(object):
         pass
 
     def parse(self, yaml_file):
-        with open(yaml_file, "r") as file:
+        with codecs.open(txt_file, "r", sys.stdin.encoding) as file:
             content = file.read()
         return yaml.load(content, Loader=yaml.BaseLoader)
 
@@ -67,10 +69,14 @@ class TextFileParser(object):
     def _check(self):
         pass
 
-    def parse(self, yaml_file):
-        with open(yaml_file, "r") as file:
-            contents = file.read().split("\n")
-        return [content for content in contents if content]
+    def parse(self, txt_file):
+        with codecs.open(txt_file, "r", sys.stdin.encoding) as file:
+            contents = []
+            for line in file:
+                line = line.strip()
+                if line:
+                    contents.append(line)
+        return contents
 
 
 csv_parser = CSVFileParser()
-- 
GitLab