Make check_xml_encodable to only validate utf-8 correctness

swift/account/server.py

@@ -32,7 +32,7 @@ from swift.common.db import AccountBroker
 from swift.common.utils import get_logger, get_param, hash_path, \
     normalize_timestamp, split_path, storage_directory
 from swift.common.constraints import ACCOUNT_LISTING_LIMIT, \
-    check_mount, check_float, check_xml_encodable
+    check_mount, check_float, check_utf8
 from swift.common.db_replicator import ReplicatorRpc
 
 
@@ -294,7 +294,7 @@ class AccountController(object):
     def __call__(self, env, start_response):
         start_time = time.time()
         req = Request(env)
-        if not check_xml_encodable(req.path_info):
+        if not check_utf8(req.path_info):
             res = HTTPPreconditionFailed(body='Invalid UTF8')
         else:
             try:

swift/common/constraints.py

@@ -108,7 +108,7 @@ def check_object_creation(req, object_name):
     if 'Content-Type' not in req.headers:
         return HTTPBadRequest(request=req, content_type='text/plain',
                     body='No content type')
-    if not check_xml_encodable(req.headers['Content-Type']):
+    if not check_utf8(req.headers['Content-Type']):
         return HTTPBadRequest(request=req, body='Invalid Content-Type',
                     content_type='text/plain')
     return check_metadata(req, 'object')
@@ -148,14 +148,15 @@ _invalid_xml = re.compile(ur'[^\x09\x0a\x0d\x20-\uD7FF\uE000-\uFFFD%s-%s]' %
                         (unichr(0x10000), unichr(0x10FFFF)))
 
 
-def check_xml_encodable(string):
+def check_utf8(string):
     """
-    Validate if a string can be encoded in xml.
+    Validate if a string is valid UTF-8.
 
     :param string: string to be validated
-    :returns: True if the string can be encoded in xml, False otherwise
+    :returns: True if the string is valid utf-8, False otherwise
     """
     try:
-        return not _invalid_xml.search(string.decode('UTF-8'))
+        string.decode('UTF-8')
+        return True
     except UnicodeDecodeError:
         return False

swift/container/server.py

@@ -33,7 +33,7 @@ from swift.common.db import ContainerBroker
 from swift.common.utils import get_logger, get_param, hash_path, \
     normalize_timestamp, storage_directory, split_path
 from swift.common.constraints import CONTAINER_LISTING_LIMIT, \
-    check_mount, check_float, check_xml_encodable
+    check_mount, check_float, check_utf8
 from swift.common.bufferedhttp import http_connect
 from swift.common.exceptions import ConnectionTimeout
 from swift.common.db_replicator import ReplicatorRpc
@@ -386,7 +386,7 @@ class ContainerController(object):
     def __call__(self, env, start_response):
         start_time = time.time()
         req = Request(env)
-        if not check_xml_encodable(req.path_info):
+        if not check_utf8(req.path_info):
             res = HTTPPreconditionFailed(body='Invalid UTF8')
         else:
             try:

swift/obj/server.py

@@ -40,7 +40,7 @@ from swift.common.utils import mkdirs, normalize_timestamp, \
     split_path, drop_buffer_cache, get_logger
 from swift.common.bufferedhttp import http_connect
 from swift.common.constraints import check_object_creation, check_mount, \
-    check_float, check_xml_encodable
+    check_float, check_utf8
 from swift.common.exceptions import ConnectionTimeout
 from swift.obj.replicator import get_hashes, invalidate_hash, \
     recalculate_hashes
@@ -557,7 +557,7 @@ class ObjectController(object):
         """WSGI Application entry point for the Swift Object Server."""
         start_time = time.time()
         req = Request(env)
-        if not check_xml_encodable(req.path_info):
+        if not check_utf8(req.path_info):
             res = HTTPPreconditionFailed(body='Invalid UTF8')
         else:
             try:

swift/proxy/server.py

@@ -36,7 +36,7 @@ from swift.common.utils import get_logger, normalize_timestamp, split_path, \
     cache_from_env
 from swift.common.bufferedhttp import http_connect
 from swift.common.constraints import check_metadata, check_object_creation, \
-    check_xml_encodable, MAX_ACCOUNT_NAME_LENGTH, MAX_CONTAINER_NAME_LENGTH, \
+    check_utf8, MAX_ACCOUNT_NAME_LENGTH, MAX_CONTAINER_NAME_LENGTH, \
     MAX_FILE_SIZE
 from swift.common.exceptions import ChunkReadTimeout, \
     ChunkWriteTimeout, ConnectionTimeout
@@ -1298,7 +1298,7 @@ class BaseApplication(object):
                 controller, path_parts = self.get_controller(req.path)
             except ValueError:
                 return HTTPNotFound(request=req)
-            if not check_xml_encodable(req.path_info):
+            if not check_utf8(req.path_info):
                 return HTTPPreconditionFailed(request=req, body='Invalid UTF8')
             if not controller:
                 return HTTPPreconditionFailed(request=req, body='Bad URL')