httputil: Fix rfc2231 for python2

Add a test that goes through parse_multipart_form_data

diff --git a/tornado/httputil.py b/tornado/httputil.py
index e6f3fc975e3ad36170c0ca688509a747b4c18175..eabe11b08a2f46e1642bd737e4e2ab60c19de2ba 100644 (file)
--- a/tornado/httputil.py
+++ b/tornado/httputil.py
@@ -891,7 +891,8 @@ def parse_response_start_line(line):
 # The original 2.7 version of this code did not correctly support some
 # combinations of semicolons and double quotes.
 # It has also been modified to support valueless parameters as seen in
-# websocket extension negotiations.
+# websocket extension negotiations, and to support non-ascii values in
+# RFC 2231/5987 format.
 
 
 def _parseparam(s):
@@ -912,8 +913,10 @@ def _parse_header(line):
 
     Return the main content-type and a dictionary of options.
 
-    >>> d = "CD: fd; foo=\"b\\\\a\\\"r\"; file*=utf-8''T%C3%A4st"
-    >>> d = _parse_header(d)[1]
+    >>> d = "form-data; foo=\"b\\\\a\\\"r\"; file*=utf-8''T%C3%A4st"
+    >>> ct, d = _parse_header(d)
+    >>> ct
+    'form-data'
     >>> d['file'] == r'T\u00e4st'.encode('ascii').decode('unicode_escape')
     True
     >>> d['foo']
@@ -928,9 +931,9 @@ def _parse_header(line):
         if i >= 0:
             name = p[:i].strip().lower()
             value = p[i + 1:].strip()
-            params.append((name, value))
+            params.append((name, native_str(value)))
     params = email.utils.decode_params(params)
-    params.pop(0) # get rid of the dummy again
+    params.pop(0)  # get rid of the dummy again
     pdict = {}
     for name, value in params:
         value = email.utils.collapse_rfc2231_value(value)

diff --git a/tornado/test/httputil_test.py b/tornado/test/httputil_test.py
index bd5832bcde40c776dd8f7a9bffa745751fe0a31f..faf7717fe1a8c569ba3d62cdc3430939409d8ddd 100644 (file)
--- a/tornado/test/httputil_test.py
+++ b/tornado/test/httputil_test.py
@@ -176,6 +176,20 @@ Foo
             self.assertEqual(file["filename"], filename)
             self.assertEqual(file["body"], b"Foo")
 
+    def test_non_ascii_filename(self):
+        data = b"""\
+--1234
+Content-Disposition: form-data; name="files"; filename="ab.txt"; filename*=UTF-8''%C3%A1b.txt
+
+Foo
+--1234--""".replace(b"\n", b"\r\n")
+        args = {}
+        files = {}
+        parse_multipart_form_data(b"1234", data, args, files)
+        file = files["files"][0]
+        self.assertEqual(file["filename"], u"áb.txt")
+        self.assertEqual(file["body"], b"Foo")
+
     def test_boundary_starts_and_ends_with_quotes(self):
         data = b'''\
 --1234