Index: Lib/pydoc.py
===================================================================
--- Lib/pydoc.py	(revision 78302)
+++ Lib/pydoc.py	(working copy)
@@ -52,7 +52,7 @@
 #     the current directory is changed with os.chdir(), an incorrect
 #     path will be displayed.
 
-import sys, imp, os, re, types, inspect, __builtin__, pkgutil
+import sys, imp, os, re, types, inspect, __builtin__, pkgutil, locale
 from repr import Repr
 from string import expandtabs, find, join, lower, split, strip, rfind, rstrip
 from traceback import extract_tb
@@ -1344,6 +1344,14 @@
     finally:
         os.unlink(filename)
 
+def encoded(text):
+    """Encode unicode text."""
+    if isinstance(text, unicode):
+        output_encoding = (getattr(sys.stdout, 'encoding', None) or
+                           locale.getpreferredencoding() or 'ascii')
+        return text.encode(output_encoding, 'replace')
+    return text
+
 def plain(text):
     """Remove boldface formatting from text."""
     return re.sub('.\b', '', text)
@@ -1352,7 +1360,7 @@
     """Page through text by feeding it to another program."""
     pipe = os.popen(cmd, 'w')
     try:
-        pipe.write(text)
+        pipe.write(encoded(text))
         pipe.close()
     except IOError:
         pass # Ignore broken pipes caused by quitting the pager program.
@@ -1362,7 +1370,7 @@
     import tempfile
     filename = tempfile.mktemp()
     file = open(filename, 'w')
-    file.write(text)
+    file.write(encoded(text))
     file.close()
     try:
         os.system(cmd + ' "' + filename + '"')
@@ -1371,6 +1379,7 @@
 
 def ttypager(text):
     """Page through text on a text terminal."""
+    text = encoded(text)
     lines = split(plain(text), '\n')
     try:
         import tty
@@ -1409,6 +1418,7 @@
 
 def plainpager(text):
     """Simply print unformatted text.  This is the ultimate fallback."""
+    text = encoded(text)
     sys.stdout.write(plain(text))
 
 def describe(thing):
Index: Lib/test/pydoc_mod.py
===================================================================
--- Lib/test/pydoc_mod.py	(revision 78302)
+++ Lib/test/pydoc_mod.py	(working copy)
@@ -16,11 +16,12 @@
     pass
 
 def doc_func():
-    """
+    u"""
     This function solves all of the world's problems:
     hunger
     lack of Python
     war
+    \xfcnicode\u2026
     """
 
 def nodoc_func():
Index: Lib/test/test_pydoc.py
===================================================================
--- Lib/test/test_pydoc.py	(revision 78302)
+++ Lib/test/test_pydoc.py	(working copy)
@@ -1,21 +1,20 @@
 import sys
 import os
-import os.path
 import difflib
+import locale
 import subprocess
-import re
 import pydoc
 import inspect
 import unittest
-import test.test_support
 from contextlib import contextmanager
 from test.test_support import (
+    run_unittest, captured_stdout,
     TESTFN, forget, rmtree, EnvironmentVarGuard, reap_children)
 
 from test import pydoc_mod
 
 expected_text_pattern = \
-"""
+u"""
 NAME
     test.pydoc_mod - This is a test module for test_pydoc
 
@@ -55,6 +54,7 @@
         hunger
         lack of Python
         war
+        \xfcnicode\u2026
 \x20\x20\x20\x20
     nodoc_func()
 
@@ -74,7 +74,7 @@
 """.strip()
 
 expected_html_pattern = \
-"""
+u"""
 <table width="100%%" cellspacing=0 cellpadding=2 border=0 summary="heading">
 <tr bgcolor="#7799ee">
 <td valign=bottom>&nbsp;<br>
@@ -138,7 +138,8 @@
 <td width="100%%"><dl><dt><a name="-doc_func"><strong>doc_func</strong></a>()</dt><dd><tt>This&nbsp;function&nbsp;solves&nbsp;all&nbsp;of&nbsp;the&nbsp;world's&nbsp;problems:<br>
 hunger<br>
 lack&nbsp;of&nbsp;Python<br>
-war</tt></dd></dl>
+war<br>
+\xfcnicode\u2026</tt></dd></dl>
  <dl><dt><a name="-nodoc_func"><strong>nodoc_func</strong></a>()</dt></dl>
 </td></tr></table><p>
 <table width="100%%" cellspacing=0 cellpadding=2 border=0 summary="section">
@@ -203,9 +204,10 @@
 
     output = doc.docmodule(module)
 
-    # cleanup the extra text formatting that pydoc preforms
-    patt = re.compile('\b.')
-    output = patt.sub('', output)
+    # go through the pydoc pager
+    with captured_stdout() as stdout:
+        pydoc.pager(output)
+    output = stdout.getvalue()
     return output.strip(), loc
 
 def print_diffs(text1, text2):
@@ -216,6 +218,10 @@
                                  tofile='got')
     print '\n' + ''.join(diffs)
 
+def encoded(text):
+    "Return encoded string"
+    return text.encode(locale.getpreferredencoding(), 'replace')
+
 
 class PyDocDocTest(unittest.TestCase):
 
@@ -236,10 +242,33 @@
         result, doc_loc = get_pydoc_text(pydoc_mod)
         expected_text = expected_text_pattern % \
                         (inspect.getabsfile(pydoc_mod), doc_loc)
+        expected_text = encoded(expected_text)
         if result != expected_text:
             print_diffs(expected_text, result)
             self.fail("outputs are not equal, see diff above")
 
+    def test_pager_unicode(self):
+        docstring = u"f\xfcr Elise"
+        with captured_stdout() as stdout:
+            pydoc.pager(docstring)
+        self.assertEqual(stdout.getvalue(), encoded(docstring))
+
+    def test_pager_bytes(self):
+        docstring = u"f\xfcr Elise"
+        for docbytes in docstring.encode('latin-1'), docstring.encode('utf-8'):
+            with captured_stdout() as stdout:
+                pydoc.pager(docbytes)
+            self.assertEqual(stdout.getvalue(), docbytes)
+
+    def test_bogus_bytestring(self):
+        # This is preserved for backward compatibility.
+        # It should not be used, because it assumes that the
+        # docstring encoding is the same as the user preferred encoding.
+        def ludwig_van_b():
+            """f\xfcr Elise"""
+        result, doc_loc = get_pydoc_text(ludwig_van_b)
+        self.assertIn(ludwig_van_b.__doc__, result)
+
     def test_not_here(self):
         missing_module = "test.i_am_not_here"
         result = run_pydoc(missing_module)
@@ -332,8 +361,7 @@
 
 
 def test_main():
-    test.test_support.run_unittest(PyDocDocTest,
-                                   TestDescriptions)
+    run_unittest(PyDocDocTest, TestDescriptions)
 
 if __name__ == "__main__":
     test_main()