Enforce ascii filenames for uploads

3 files changed, 15 insertions, 1 deletions

diff --git a/synapse/rest/media/v1/base_resource.py b/synapse/rest/media/v1/base_resource.py
index 04410ab827..1b7517e2f0 100644
--- a/synapse/rest/media/v1/base_resource.py
+++ b/synapse/rest/media/v1/base_resource.py
@@ -27,6 +27,7 @@ from twisted.web.resource import Resource
 from twisted.protocols.basic import FileSender
 
 from synapse.util.async import ObservableDeferred
+from synapse.util.stringutils import is_ascii
 
 import os
 
@@ -135,6 +136,8 @@ class BaseMediaResource(Resource):
             if content_disposition:
                 _, params = cgi.parse_header(content_disposition[0],)
                 upload_name = params.get("filename", None)
+                if upload_name and not is_ascii(upload_name):
+                    upload_name = None
             else:
                 upload_name = None
 
diff --git a/synapse/rest/media/v1/upload_resource.py b/synapse/rest/media/v1/upload_resource.py
index 92e855a448..cdd1d44e07 100644
--- a/synapse/rest/media/v1/upload_resource.py
+++ b/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
+from synapse.util.stringutils import random_string, is_ascii
 from synapse.api.errors import SynapseError
 
 from twisted.web.server import NOT_DONE_YET
@@ -87,6 +87,8 @@ 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")
 
         headers = request.requestHeaders
 
diff --git a/synapse/util/stringutils.py b/synapse/util/stringutils.py
index 52e66beaee..7a1e96af37 100644
--- a/synapse/util/stringutils.py
+++ b/synapse/util/stringutils.py
@@ -33,3 +33,12 @@ def random_string_with_symbols(length):
     return ''.join(
         random.choice(_string_with_symbols) for _ in xrange(length)
     )
+
+
+def is_ascii(s):
+    try:
+        s.encode("ascii")
+    except UnicodeDecodeError:
+        return False
+    else:
+        return True