Update ycmd, and use JoinLinesAsUnicode in GetUnsavedAndCurrentBufferData().

This function is invoked on every kepress.  On a large file (15+k loc),
using this new function takes kepress latency down to <10ms from ~100ms.

python/ycm/tests/vimsupport_test.py

@@ -1407,7 +1407,7 @@ def GetUnsavedAndCurrentBufferData_EncodedUnicodeCharsInBuffers_test( *args ):
   mock_buffer = MagicMock()
   mock_buffer.name = os.path.realpath( 'filename' )
   mock_buffer.number = 1
-  mock_buffer.__iter__.return_value = [ u'abc', ToBytes( u'fДa' ) ]
+  mock_buffer.__iter__.return_value = [ ToBytes ( u'abc' ), ToBytes( u'fДa' ) ]
 
   with patch( 'vim.buffers', [ mock_buffer ] ):
     assert_that( vimsupport.GetUnsavedAndCurrentBufferData(),

python/ycm/vimsupport.py

@@ -30,7 +30,7 @@ import tempfile
 import json
 import re
 from collections import defaultdict
-from ycmd.utils import ToUnicode, ToBytes
+from ycmd.utils import ToUnicode, ToBytes, JoinLinesAsUnicode
 from ycmd import user_options_store
 
 BUFFER_COMMAND_MAP = { 'same-buffer'      : 'edit',
@@ -125,7 +125,7 @@ def GetUnsavedAndCurrentBufferData():
 
     buffers_data[ GetBufferFilepath( buffer_object ) ] = {
       # Add a newline to match what gets saved to disk. See #1455 for details.
-      'contents': '\n'.join( ToUnicode( x ) for x in buffer_object ) + '\n',
+      'contents': JoinLinesAsUnicode( buffer_object ) + '\n',
       'filetypes': FiletypesForBuffer( buffer_object )
     }
 

third_party/ycmd

@@ -1 +1 @@
-Subproject commit 363506449fe18337b6aabbf29d8a14720483ccf6
+Subproject commit f3232ce5180753822d839c80e9b9ea4e33e89d4c