Checkpoint. Manipulated things so that string literals are always

unicode, and a few other compensating changes, e.g. str <- unicode,
chr <- unichr, and repr() of a unicode string no longer starts
with 'u'.  Lots of unit tests are broken, but some basic things
work, in particular distutils works so the extensions can be built,
and test_builtin.py works.

Author: gvanrossum

Include/pydebug.h

@@ -14,7 +14,6 @@ PyAPI_DATA(int) Py_NoSiteFlag;
 PyAPI_DATA(int) Py_UseClassExceptionsFlag;
 PyAPI_DATA(int) Py_FrozenFlag;
 PyAPI_DATA(int) Py_TabcheckFlag;
-PyAPI_DATA(int) Py_UnicodeFlag;
 PyAPI_DATA(int) Py_IgnoreEnvironmentFlag;
 PyAPI_DATA(int) Py_DivisionWarningFlag;
 

Lib/distutils/ccompiler.py

@@ -168,7 +168,7 @@ class (via the 'executables' class attribute), but most will have:
     # set_executables ()
 
     def set_executable(self, key, value):
-        if type(value) is StringType:
+        if isinstance(value, basestring):
             setattr(self, key, split_quoted(value))
         else:
             setattr(self, key, value)
@@ -193,8 +193,8 @@ def _check_macro_definitions (self, definitions):
             if not (type (defn) is TupleType and
                     (len (defn) == 1 or
                      (len (defn) == 2 and
-                      (type (defn[1]) is StringType or defn[1] is None))) and
-                    type (defn[0]) is StringType):
+                      (isinstance (defn[1], basestring) or defn[1] is None))) and
+                    isinstance (defn[0], basestring)):
                 raise TypeError, \
                       ("invalid macro definition '%s': " % defn) + \
                       "must be tuple (string,), (string, string), or " + \
@@ -344,7 +344,7 @@ def _setup_compile(self, outdir, macros, incdirs, sources, depends,
         """
         if outdir is None:
             outdir = self.output_dir
-        elif type(outdir) is not StringType:
+        elif not isinstance(outdir, basestring):
             raise TypeError, "'output_dir' must be a string or None"
 
         if macros is None:
@@ -442,7 +442,7 @@ def _fix_compile_args (self, output_dir, macros, include_dirs):
         """
         if output_dir is None:
             output_dir = self.output_dir
-        elif type (output_dir) is not StringType:
+        elif not isinstance(output_dir, basestring):
             raise TypeError, "'output_dir' must be a string or None"
 
         if macros is None:
@@ -527,7 +527,7 @@ def _fix_object_args (self, objects, output_dir):
 
         if output_dir is None:
             output_dir = self.output_dir
-        elif type (output_dir) is not StringType:
+        elif not isinstance(output_dir, basestring):
             raise TypeError, "'output_dir' must be a string or None"
 
         return (objects, output_dir)

Lib/distutils/cmd.py

@@ -222,7 +222,7 @@ def _ensure_stringlike (self, option, what, default=None):
         if val is None:
             setattr(self, option, default)
             return default
-        elif type(val) is not StringType:
+        elif not isinstance(val, basestring):
             raise DistutilsOptionError, \
                   "'%s' must be a %s (got `%s`)" % (option, what, val)
         return val
@@ -242,12 +242,11 @@ def ensure_string_list (self, option):
         val = getattr(self, option)
         if val is None:
             return
-        elif type(val) is StringType:
+        elif isinstance(val, basestring):
             setattr(self, option, re.split(r',\s*|\s+', val))
         else:
             if type(val) is ListType:
-                types = map(type, val)
-                ok = (types == [StringType] * len(val))
+                ok = all(isinstance(v, basestring) for v in val)
             else:
                 ok = 0
 
@@ -421,7 +420,7 @@ def make_file (self, infiles, outfile, func, args,
 
 
         # Allow 'infiles' to be a single string
-        if type(infiles) is StringType:
+        if isinstance(infiles, basestring):
             infiles = (infiles,)
         elif type(infiles) not in (ListType, TupleType):
             raise TypeError, \

Lib/distutils/command/build_clib.py

@@ -92,7 +92,7 @@ def finalize_options (self):
 
         if self.include_dirs is None:
             self.include_dirs = self.distribution.include_dirs or []
-        if type(self.include_dirs) is StringType:
+        if isinstance(self.include_dirs, basestring):
             self.include_dirs = self.include_dirs.split(os.pathsep)
 
         # XXX same as for build_ext -- what about 'self.define' and
@@ -147,7 +147,7 @@ def check_library_list (self, libraries):
                 raise DistutilsSetupError, \
                       "each element of 'libraries' must a 2-tuple"
 
-            if type(lib[0]) is not StringType:
+            if isinstance(lib[0], basestring) StringType:
                 raise DistutilsSetupError, \
                       "first element of each tuple in 'libraries' " + \
                       "must be a string (the library name)"

Lib/distutils/command/build_ext.py

@@ -137,7 +137,7 @@ def finalize_options (self):
         plat_py_include = sysconfig.get_python_inc(plat_specific=1)
         if self.include_dirs is None:
             self.include_dirs = self.distribution.include_dirs or []
-        if type(self.include_dirs) is StringType:
+        if isinstance(self.include_dirs, basestring):
             self.include_dirs = self.include_dirs.split(os.pathsep)
 
         # Put the Python "system" include dir at the end, so that
@@ -146,7 +146,7 @@ def finalize_options (self):
         if plat_py_include != py_include:
             self.include_dirs.append(plat_py_include)
 
-        if type(self.libraries) is StringType:
+        if isinstance(self.libraries, basestring):
             self.libraries = [self.libraries]
 
         # Life is easier if we're not forever checking for None, so
@@ -155,12 +155,12 @@ def finalize_options (self):
             self.libraries = []
         if self.library_dirs is None:
             self.library_dirs = []
-        elif type(self.library_dirs) is StringType:
+        elif isinstance(self.library_dirs, basestring):
             self.library_dirs = self.library_dirs.split(os.pathsep)
 
         if self.rpath is None:
             self.rpath = []
-        elif type(self.rpath) is StringType:
+        elif isinstance(self.rpath, basestring):
             self.rpath = self.rpath.split(os.pathsep)
 
         # for extensions under windows use different directories
@@ -321,7 +321,7 @@ def check_extensions_list (self, extensions):
                       ("each element of 'ext_modules' option must be an "
                        "Extension instance or 2-tuple")
 
-            if not (type(ext_name) is StringType and
+            if not (isinstance(ext_name, basestring) and
                     extension_name_re.match(ext_name)):
                 raise DistutilsSetupError, \
                       ("first element of each tuple in 'ext_modules' "

Lib/distutils/command/build_py.py

@@ -361,7 +361,7 @@ def get_outputs (self, include_bytecode=1):
 
 
     def build_module (self, module, module_file, package):
-        if type(package) is StringType:
+        if isinstance(package, basestring):
             package = package.split('.')
         elif type(package) not in (ListType, TupleType):
             raise TypeError, \

Lib/distutils/command/config.py

@@ -73,17 +73,17 @@ def initialize_options (self):
     def finalize_options (self):
         if self.include_dirs is None:
             self.include_dirs = self.distribution.include_dirs or []
-        elif type(self.include_dirs) is StringType:
+        elif isinstance(self.include_dirs, basestring):
             self.include_dirs = self.include_dirs.split(os.pathsep)
 
         if self.libraries is None:
             self.libraries = []
-        elif type(self.libraries) is StringType:
+        elif isinstance(self.libraries, basestring):
             self.libraries = [self.libraries]
 
         if self.library_dirs is None:
             self.library_dirs = []
-        elif type(self.library_dirs) is StringType:
+        elif isinstance(self.library_dirs, basestring):
             self.library_dirs = self.library_dirs.split(os.pathsep)
 
 
@@ -212,7 +212,7 @@ def search_cpp (self, pattern, body=None,
         self._check_compiler()
         (src, out) = self._preprocess(body, headers, include_dirs, lang)
 
-        if type(pattern) is StringType:
+        if isinstance(pattern, basestring):
             pattern = re.compile(pattern)
 
         file = open(out)

Lib/distutils/command/install.py

@@ -463,7 +463,7 @@ def handle_extra_path (self):
             self.extra_path = self.distribution.extra_path
 
         if self.extra_path is not None:
-            if type(self.extra_path) is StringType:
+            if isinstance(self.extra_path, basestring):
                 self.extra_path = self.extra_path.split(',')
 
             if len(self.extra_path) == 1:

Lib/distutils/command/install_data.py

@@ -10,7 +10,6 @@
 __revision__ = "$Id$"
 
 import os
-from types import StringType
 from distutils.core import Command
 from distutils.util import change_root, convert_path
 
@@ -48,7 +47,7 @@ def finalize_options (self):
     def run (self):
         self.mkpath(self.install_dir)
         for f in self.data_files:
-            if type(f) is StringType:
+            if isinstance(f, basestring):
                 # it's a simple file, so copy it
                 f = convert_path(f)
                 if self.warn_dir:

Lib/distutils/dir_util.py

@@ -31,7 +31,7 @@ def mkpath (name, mode=0777, verbose=0, dry_run=0):
     global _path_created
 
     # Detect a common bug -- name is None
-    if not isinstance(name, StringTypes):
+    if not isinstance(name, basestring):
         raise DistutilsInternalError, \
               "mkpath: 'name' must be a string (got %r)" % (name,)
 

Lib/distutils/dist.py

@@ -598,13 +598,13 @@ def finalize_options (self):
 
         keywords = self.metadata.keywords
         if keywords is not None:
-            if type(keywords) is StringType:
+            if isinstance(keywords, basestring):
                 keywordlist = keywords.split(',')
                 self.metadata.keywords = [x.strip() for x in keywordlist]
 
         platforms = self.metadata.platforms
         if platforms is not None:
-            if type(platforms) is StringType:
+            if isinstance(platforms, basestring):
                 platformlist = platforms.split(',')
                 self.metadata.platforms = [x.strip() for x in platformlist]
 
@@ -906,7 +906,7 @@ def _set_command_options (self, command_obj, option_dict=None):
                 neg_opt = {}
 
             try:
-                is_string = type(value) is StringType
+                is_string = isinstance(value, basestring)
                 if option in neg_opt and is_string:
                     setattr(command_obj, neg_opt[option], not strtobool(value))
                 elif option in bool_opts and is_string:

Lib/distutils/extension.py

@@ -103,9 +103,9 @@ def __init__ (self, name, sources,
                   language=None,
                   **kw                      # To catch unknown keywords
                  ):
-        assert type(name) is StringType, "'name' must be a string"
+        assert isinstance(name, basestring), "'name' must be a string"
         assert (type(sources) is ListType and
-                map(type, sources) == [StringType]*len(sources)), \
+                all(isinstance(v, basestring) for v in sources)), \
                 "'sources' must be a list of strings"
 
         self.name = name

Lib/distutils/fancy_getopt.py

@@ -166,13 +166,13 @@ def _grok_option_table (self):
                 raise ValueError, "invalid option tuple: %r" % (option,)
 
             # Type- and value-check the option names
-            if type(long) is not StringType or len(long) < 2:
+            if not isinstance(long, basestring) or len(long) < 2:
                 raise DistutilsGetoptError, \
                       ("invalid long option '%s': "
                        "must be a string of length >= 2") % long
 
             if (not ((short is None) or
-                     (type(short) is StringType and len(short) == 1))):
+                     (isinstance(short, basestring) and len(short) == 1))):
                 raise DistutilsGetoptError, \
                       ("invalid short option '%s': "
                        "must a single character or None") % short

Lib/distutils/filelist.py

@@ -333,7 +333,7 @@ def translate_pattern (pattern, anchor=1, prefix=None, is_regex=0):
     or just returned as-is (assumes it's a regex object).
     """
     if is_regex:
-        if type(pattern) is StringType:
+        if isinstance(pattern, basestring):
             return re.compile(pattern)
         else:
             return pattern

Lib/distutils/unixccompiler.py

@@ -16,7 +16,7 @@
 __revision__ = "$Id$"
 
 import os, sys
-from types import StringType, NoneType
+from types import NoneType
 from copy import copy
 
 from distutils import sysconfig
@@ -212,7 +212,7 @@ def link(self, target_desc, objects,
 
         lib_opts = gen_lib_options(self, library_dirs, runtime_library_dirs,
                                    libraries)
-        if type(output_dir) not in (StringType, NoneType):
+        if not isinstance(output_dir, (basestring, NoneType)):
             raise TypeError, "'output_dir' must be a string or None"
         if output_dir is not None:
             output_filename = os.path.join(output_dir, output_filename)

Lib/locale.py

@@ -470,7 +470,7 @@ def setlocale(category, locale=None):
         category may be given as one of the LC_* values.
 
     """
-    if locale and type(locale) is not type(""):
+    if locale and not isinstance(locale, basestring):
         # convert to string
         locale = normalize(_build_localename(locale))
     return _setlocale(category, locale)

Lib/os.py

@@ -733,8 +733,8 @@ def urandom(n):
             _urandomfd = open("/dev/urandom", O_RDONLY)
         except (OSError, IOError):
             raise NotImplementedError("/dev/urandom (or equivalent) not found")
-        bytes = ""
-        while len(bytes) < n:
-            bytes += read(_urandomfd, n - len(bytes))
+        bs = b""
+        while len(bs) < n:
+            bs += read(_urandomfd, n - len(bs))
         close(_urandomfd)
-        return bytes
+        return bs

Lib/sre_compile.py

@@ -470,18 +470,8 @@ def _compile_info(code, pattern, flags):
         _compile_charset(charset, flags, code)
     code[skip] = len(code) - skip
 
-try:
-    unicode
-except NameError:
-    STRING_TYPES = (type(""),)
-else:
-    STRING_TYPES = (type(""), type(unicode("")))
-
 def isstring(obj):
-    for tp in STRING_TYPES:
-        if isinstance(obj, tp):
-            return 1
-    return 0
+    return isinstance(obj, basestring)
 
 def _code(p, flags):