Better handling of unicode in snippet definitions and names

plugin/UltiSnips/Util.py

@@ -1,8 +1,14 @@
 #!/usr/bin/env python
 # encoding: utf-8
 
-import vim
 import os
+import types
+import vim
+
+def as_utf8(s):
+    if not isinstance(s, types.UnicodeType):
+        s = s.decode("utf-8")
+    return s.encode("utf-8")
 
 def vim_string(inp):
     """ Creates a vim-friendly string from a group of
@@ -10,14 +16,15 @@ def vim_string(inp):
     """
     def conv(obj):
         if isinstance(obj, list):
-            return u'[' + u",".join([conv(o) for o in obj]) + u']'
+            rv = u'[' + u','.join(conv(o) for o in obj) + u']'
         elif isinstance(obj, dict):
-            return u'{' + u','.join([
+            rv = u'{' + u','.join([
                 u"%s:%s" % (conv(key), conv(value))
                 for key, value in obj.iteritems()]) + u'}'
         else:
-            return u'"%s"' % str(obj).replace(u'"', u'\\"')
-    return conv(inp)
+            rv = u'"%s"' % str(obj).decode("utf-8").replace(u'"', u'\\"')
+        return rv
+    return conv(inp).encode("utf-8")
 
 class IndentUtil(object):
     """ Utility class for dealing properly with indentation. """

plugin/UltiSnips/__init__.py

@@ -14,7 +14,7 @@ import vim
 from UltiSnips.Geometry import Position
 from UltiSnips.TextObjects import *
 from UltiSnips.Buffer import VimBuffer
-from UltiSnips.Util import IndentUtil, vim_string
+from UltiSnips.Util import IndentUtil, vim_string, as_utf8
 from UltiSnips.Langmap import LangMapTranslator
 
 # The following lines silence DeprecationWarnings. They are raised
@@ -266,9 +266,9 @@ class Snippet(object):
     _TABS = re.compile(r"^\t*")
 
     def __init__(self, trigger, value, descr, options, globals):
-        self._t = trigger
-        self._v = value
-        self._d = descr
+        self._t = as_utf8(trigger)
+        self._v = as_utf8(value)
+        self._d = as_utf8(descr)
         self._opts = options
         self._matched = ""
         self._last_re = None

plugin/UltiSnips/debug.py

@@ -3,9 +3,13 @@
 
 __all__ = [ "debug" ]
 
+import types
+
 def debug(s):
+    if not isinstance(s, types.UnicodeType):
+        s = s.decode("utf-8")
     f = open("/tmp/file.txt","a")
-    f.write(s+'\n')
+    f.write(s.encode("utf-8")+'\n')
     f.close()
 
 

test.py

@@ -1912,6 +1912,16 @@ class Snippet_With_DoubleQuote(_VimTest):
     keys = 'te"st' + EX
     wanted = "Expand me\"!"
 
+class Snippet_With_Umlauts_List(_VimTest):
+    snippets = _snip_quote(u'ü')
+    keys = 'te' + LS + "2\n"
+    wanted = "Expand meü!"
+
+class Snippet_With_Umlauts(_VimTest):
+    snippets = _snip_quote(u'ü')
+    keys = u'teüst' + EX
+    wanted = "Expand meü!"
+
 class Snippet_With_DoubleQuote_List(_VimTest):
     snippets = _snip_quote('"')
     keys = "te" + LS + "2\n"