Remove the u() unicode function

jedi/_compatibility.py

@@ -134,18 +134,7 @@ class ImplicitNSInfo(object):
         self.paths = paths
 
 
-def u(string, errors='strict'):
-    """Cast to unicode DAMMIT!
-    Written because Python2 repr always implicitly casts to a string, so we
-    have to cast back to a unicode (and we now that we always deal with valid
-    unicode, because we check that in the beginning).
-    """
-    if isinstance(string, bytes):
-        return str(string, encoding='UTF-8', errors=errors)
-    return string
-
-
-def cast_path(obj):
+def cast_path(string):
     """
     Take a bytes or str path and cast it to unicode.
 
@@ -156,7 +145,9 @@ def cast_path(obj):
     Since this just really complicates everything and Python 2.7 will be EOL
     soon anyway, just go with always strings.
     """
-    return u(obj, errors='replace')
+    if isinstance(string, bytes):
+        return str(string, encoding='UTF-8', errors='replace')
+    return string
 
 
 def pickle_load(file):

jedi/api/helpers.py

@@ -11,7 +11,6 @@ from inspect import Parameter
 from parso.python.parser import Parser
 from parso.python import tree
 
-from jedi._compatibility import u
 from jedi.inference.base_value import NO_VALUES
 from jedi.inference.syntax_tree import infer_atom
 from jedi.inference.helpers import infer_call_of_leaf
@@ -85,18 +84,18 @@ def _get_code_for_stack(code_lines, leaf, position):
         # If we're not on a comment simply get the previous leaf and proceed.
         leaf = leaf.get_previous_leaf()
         if leaf is None:
-            return u('')  # At the beginning of the file.
+            return ''  # At the beginning of the file.
 
     is_after_newline = leaf.type == 'newline'
     while leaf.type == 'newline':
         leaf = leaf.get_previous_leaf()
         if leaf is None:
-            return u('')
+            return ''
 
     if leaf.type == 'error_leaf' or leaf.type == 'string':
         if leaf.start_pos[0] < position[0]:
             # On a different line, we just begin anew.
-            return u('')
+            return ''
 
         # Error leafs cannot be parsed, completion in strings is also
         # impossible.
@@ -112,7 +111,7 @@ def _get_code_for_stack(code_lines, leaf, position):
             if user_stmt.start_pos[1] > position[1]:
                 # This means that it's actually a dedent and that means that we
                 # start without value (part of a suite).
-                return u('')
+                return ''
 
         # This is basically getting the relevant lines.
         return _get_code(code_lines, user_stmt.get_start_pos_of_prefix(), position)

jedi/debug.py

@@ -2,8 +2,6 @@ import os
 import time
 from contextlib import contextmanager
 
-from jedi._compatibility import u
-
 _inited = False
 
 
@@ -104,7 +102,7 @@ def dbg(message, *args, color='GREEN'):
     if debug_function and enable_notice:
         i = ' ' * _debug_indent
         _lazy_colorama_init()
-        debug_function(color, i + 'dbg: ' + message % tuple(u(repr(a)) for a in args))
+        debug_function(color, i + 'dbg: ' + message % tuple(repr(a) for a in args))
 
 
 def warning(message, *args, **kwargs):
@@ -114,7 +112,7 @@ def warning(message, *args, **kwargs):
     if debug_function and enable_warning:
         i = ' ' * _debug_indent
         if format:
-            message = message % tuple(u(repr(a)) for a in args)
+            message = message % tuple(repr(a) for a in args)
         debug_function('RED', i + 'warning: ' + message)
 
 

jedi/inference/context.py

@@ -438,13 +438,12 @@ def get_global_filters(context, until_position, origin_scope):
     For global name lookups. The filters will handle name resolution
     themselves, but here we gather possible filters downwards.
 
-    >>> from jedi._compatibility import u
     >>> from jedi import Script
-    >>> script = Script(u('''
+    >>> script = Script('''
     ... x = ['a', 'b', 'c']
     ... def func():
     ...     y = None
-    ... '''))
+    ... ''')
     >>> module_node = script._module_node
     >>> scope = next(module_node.iter_funcdefs())
     >>> scope

jedi/inference/docstrings.py

@@ -21,7 +21,6 @@ from textwrap import dedent
 
 from parso import parse, ParserSyntaxError
 
-from jedi._compatibility import u
 from jedi import debug
 from jedi.common import indent_block
 from jedi.inference.cache import inference_state_method_cache
@@ -184,7 +183,7 @@ def _strip_rst_role(type_str):
 
 
 def _infer_for_statement_string(module_context, string):
-    code = dedent(u("""
+    code = dedent("""
     def pseudo_docstring_stuff():
         '''
         Create a pseudo function for docstring statements.
@@ -192,7 +191,7 @@ def _infer_for_statement_string(module_context, string):
         is still a function.
         '''
     {}
-    """))
+    """)
     if string is None:
         return []
 

scripts/diff_parser_profile.py

@@ -18,7 +18,6 @@ from docopt import docopt
 from jedi.parser.python import load_grammar
 from jedi.parser.diff import DiffParser
 from jedi.parser.python import ParserWithRecovery
-from jedi._compatibility import u
 from jedi.common import splitlines
 import jedi
 
@@ -37,14 +36,15 @@ def main(args):
     with open(args['<file>']) as f:
         code = f.read()
     grammar = load_grammar()
-    parser = ParserWithRecovery(grammar, u(code))
+    parser = ParserWithRecovery(grammar, code)
     # Make sure used_names is loaded
     parser.module.used_names
 
-    code =  code + '\na\n'  # Add something so the diff parser needs to run.
+    code = code + '\na\n'  # Add something so the diff parser needs to run.
     lines = splitlines(code, keepends=True)
     cProfile.runctx('run(parser, lines)', globals(), locals(), sort=args['-s'])
 
+
 if __name__ == '__main__':
     args = docopt(__doc__)
     main(args)

test/test_api/test_unicode.py

@@ -1,7 +1,6 @@
 """
 All character set and unicode related tests.
 """
-from jedi._compatibility import u
 from jedi import Project
 
 
@@ -11,12 +10,12 @@ def test_unicode_script(Script):
     assert len(completions)
     assert type(completions[0].description) is str
 
-    s = u("author='öä'; author")
+    s = "author='öä'; author"
     completions = Script(s).complete()
     x = completions[0].description
     assert type(x) is str
 
-    s = u("#-*- coding: iso-8859-1 -*-\nauthor='öä'; author")
+    s = "#-*- coding: iso-8859-1 -*-\nauthor='öä'; author"
     s = s.encode('latin-1')
     completions = Script(s).complete()
     assert type(completions[0].description) is str
@@ -24,21 +23,21 @@ def test_unicode_script(Script):
 
 def test_unicode_attribute(Script):
     """ github jedi-vim issue #94 """
-    s1 = u('#-*- coding: utf-8 -*-\nclass Person():\n'
-           '    name = "e"\n\nPerson().name.')
+    s1 = ('#-*- coding: utf-8 -*-\nclass Person():\n'
+          '    name = "e"\n\nPerson().name.')
     completions1 = Script(s1).complete()
     assert 'strip' in [c.name for c in completions1]
-    s2 = u('#-*- coding: utf-8 -*-\nclass Person():\n'
-           '    name = "é"\n\nPerson().name.')
+    s2 = ('#-*- coding: utf-8 -*-\nclass Person():\n'
+          '    name = "é"\n\nPerson().name.')
     completions2 = Script(s2).complete()
     assert 'strip' in [c.name for c in completions2]
 
 
 def test_multibyte_script(Script):
     """ `jedi.Script` must accept multi-byte string source. """
-    code = u("import datetime; datetime.d")
-    comment = u("# multi-byte comment あいうえおä")
-    s = (u('%s\n%s') % (code, comment))
+    code = "import datetime; datetime.d"
+    comment = "# multi-byte comment あいうえおä"
+    s = ('%s\n%s') % (code, comment)
     assert len(Script(s).complete(1, len(code)))
 
 
@@ -63,7 +62,7 @@ def test_complete_at_zero(Script):
 def test_wrong_encoding(Script, tmpdir):
     x = tmpdir.join('x.py')
     # Use both latin-1 and utf-8 (a really broken file).
-    x.write_binary(u'foobar = 1\nä'.encode('latin-1') + u'ä'.encode('utf-8'))
+    x.write_binary('foobar = 1\nä'.encode('latin-1') + 'ä'.encode('utf-8'))
 
     project = Project('.', sys_path=[tmpdir.strpath])
     c, = Script('import x; x.foo', project=project).complete()