1 files changed, 31 insertions, 2 deletions
diff --git a/synapse/util/threepids.py b/synapse/util/threepids.py
index 63f955acff..eb005007d5 100644
--- a/synapse/util/threepids.py
+++ b/synapse/util/threepids.py
@@ -1,4 +1,3 @@
-# -*- coding: utf-8 -*-
 # Copyright 2018 New Vector Ltd
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
@@ -19,8 +18,18 @@ import re
 logger = logging.getLogger(__name__)
 
 
+# it's unclear what the maximum length of an email address is. RFC3696 (as corrected
+# by errata) says:
+#    the upper limit on address lengths should normally be considered to be 254.
+#
+# In practice, mail servers appear to be more tolerant and allow 400 characters
+# or so. Let's allow 500, which should be plenty for everyone.
+#
+MAX_EMAIL_ADDRESS_LENGTH = 500
+
+
 async def check_3pid_allowed(hs, medium, address):
-    """Checks whether a given 3PID is allowed to be used on this HS
+    """Checks whether a given format of 3PID is allowed to be used on this HS
 
     Args:
         hs (synapse.server.HomeServer): server
@@ -98,3 +107,23 @@ def canonicalise_email(address: str) -> str:
         raise ValueError("Unable to parse email address")
 
     return parts[0].casefold() + "@" + parts[1].lower()
+
+
+def validate_email(address: str) -> str:
+    """Does some basic validation on an email address.
+
+    Returns the canonicalised email, as returned by `canonicalise_email`.
+
+    Raises a ValueError if the email is invalid.
+    """
+    # First we try canonicalising in case that fails
+    address = canonicalise_email(address)
+
+    # Email addresses have to be at least 3 characters.
+    if len(address) < 3:
+        raise ValueError("Unable to parse email address")
+
+    if len(address) > MAX_EMAIL_ADDRESS_LENGTH:
+        raise ValueError("Unable to parse email address")
+
+    return address