Allow non-ascii filenames for attachments

synapse/rest/media/v1/base_resource.py

@@ -33,6 +33,7 @@ import os
 
 import cgi
 import logging
+import urllib
 
 logger = logging.getLogger(__name__)
 
@@ -181,10 +182,18 @@ class BaseMediaResource(Resource):
         if os.path.isfile(file_path):
             request.setHeader(b"Content-Type", media_type.encode("UTF-8"))
             if upload_name:
-                request.setHeader(
-                    b"Content-Disposition",
-                    b"inline; filename=%s" % (upload_name.encode("utf-8"),),
-                )
+                if is_ascii(upload_name):
+                    request.setHeader(
+                        b"Content-Disposition",
+                        b"inline; filename=%s" % (upload_name.encode("utf-8"),),
+                    )
+                else:
+                    request.setHeader(
+                        b"Content-Disposition",
+                        b"inline; filename*=utf-8''%s" % (
+                            urllib.quote(upload_name.encode("utf-8")),
+                        ),
+                    )
 
             # cache for at least a day.
             # XXX: we might want to turn this off for data we don't want to

synapse/rest/media/v1/upload_resource.py

@@ -15,7 +15,7 @@
 
 from synapse.http.server import respond_with_json, request_handler
 
-from synapse.util.stringutils import random_string, is_ascii
+from synapse.util.stringutils import random_string
 from synapse.api.errors import SynapseError
 
 from twisted.web.server import NOT_DONE_YET
@@ -86,9 +86,7 @@ class UploadResource(BaseMediaResource):
 
         upload_name = request.args.get("filename", None)
         if upload_name:
-            upload_name = upload_name[0]
-            if upload_name and not is_ascii(upload_name):
-                raise SynapseError(400, "filename must be ascii")
+            upload_name = upload_name[0].decode('UTF-8')
 
         headers = request.requestHeaders
 

synapse/util/stringutils.py

@@ -38,6 +38,8 @@ def random_string_with_symbols(length):
 def is_ascii(s):
     try:
         s.encode("ascii")
+    except UnicodeEncodeError:
+        return False
     except UnicodeDecodeError:
         return False
     else: