use UTF-8 as the default encoding for multipart text parts

aio-libs · Dec 16, 2016 · e61fab2 · e61fab2
1 parent ee4773a
commit e61fab2
Show file tree

Hide file tree

Showing 2 changed files with 10 additions and 1 deletion.
diff --git a/aiohttp/multipart.py b/aiohttp/multipart.py
@@ -394,7 +394,9 @@ def text(self, *, encoding=None):
         :rtype: str
         """
         data = yield from self.read(decode=True)
-        encoding = encoding or self.get_charset(default='latin1')
+        # see https://www.w3.org/TR/html5/forms.html#multipart/form-data-encoding-algorithm
+        # and https://dvcs.w3.org/hg/xhr/raw-file/tip/Overview.html#dom-xmlhttprequest-send
+        encoding = encoding or self.get_charset(default='utf-8')
         return data.decode(encoding)
 
     @asyncio.coroutine

diff --git a/tests/test_multipart.py b/tests/test_multipart.py
@@ -361,6 +361,13 @@ def test_read_text(self):
         result = yield from obj.text()
         self.assertEqual('Hello, world!', result)
 
+    def test_read_text_default_encoding(self):
+        obj = aiohttp.multipart.BodyPartReader(
+            self.boundary, {},
+            Stream('Привет, Мир!\r\n--:--'.encode('utf-8')))
+        result = yield from obj.text()
+        self.assertEqual('Привет, Мир!', result)
+
     def test_read_text_encoding(self):
         obj = aiohttp.multipart.BodyPartReader(
             self.boundary, {},