python-3.6.zip added from Github

README.cosmo contains the necessary links.

third_party/python/Tools/unicode/genwincodec.py

@@ -0,0 +1,61 @@
+"""This script generates a Python codec module from a Windows Code Page.
+
+It uses the function MultiByteToWideChar to generate a decoding table.
+"""
+
+import ctypes
+from ctypes import wintypes
+from gencodec import codegen
+import unicodedata
+
+def genwinmap(codepage):
+    MultiByteToWideChar = ctypes.windll.kernel32.MultiByteToWideChar
+    MultiByteToWideChar.argtypes = [wintypes.UINT, wintypes.DWORD,
+                                    wintypes.LPCSTR, ctypes.c_int,
+                                    wintypes.LPWSTR, ctypes.c_int]
+    MultiByteToWideChar.restype = ctypes.c_int
+
+    enc2uni = {}
+
+    for i in list(range(32)) + [127]:
+        enc2uni[i] = (i, 'CONTROL CHARACTER')
+
+    for i in range(256):
+        buf = ctypes.create_unicode_buffer(2)
+        ret = MultiByteToWideChar(
+            codepage, 0,
+            bytes([i]), 1,
+            buf, 2)
+        assert ret == 1, "invalid code page"
+        assert buf[1] == '\x00'
+        try:
+            name = unicodedata.name(buf[0])
+        except ValueError:
+            try:
+                name = enc2uni[i][1]
+            except KeyError:
+                name = ''
+
+        enc2uni[i] = (ord(buf[0]), name)
+
+    return enc2uni
+
+def genwincodec(codepage):
+    import platform
+    map = genwinmap(codepage)
+    encodingname = 'cp%d' % codepage
+    code = codegen("", map, encodingname)
+    # Replace first lines with our own docstring
+    code = '''\
+"""Python Character Mapping Codec %s generated on Windows:
+%s with the command:
+  python Tools/unicode/genwincodec.py %s
+"""#"
+''' % (encodingname, ' '.join(platform.win32_ver()), codepage
+      ) + code.split('"""#"', 1)[1]
+
+    print(code)
+
+if __name__ == '__main__':
+    import sys
+    genwincodec(int(sys.argv[1]))