#1065986: Make pydoc handle unicode strings.

Patch by Akira Kitada.
This commit is contained in:
R David Murray 2014-01-05 12:35:59 -05:00
parent 1d2ef64df6
commit 984f630f0a
3 changed files with 137 additions and 14 deletions

View File

@ -81,6 +81,7 @@ def pathdirs():
def getdoc(object):
"""Get the doc string or comments for an object."""
result = inspect.getdoc(object) or inspect.getcomments(object)
result = _encode(result)
return result and re.sub('^ *\n', '', rstrip(result)) or ''
def splitdoc(doc):
@ -182,6 +183,34 @@ def classify_class_attrs(object):
return name, kind, cls, value
return map(fixup, inspect.classify_class_attrs(object))
# ----------------------------------------------------- Unicode support helpers
try:
_unicode = unicode
except NameError:
# If Python is built without Unicode support, the unicode type
# will not exist. Fake one that nothing will match, and make
# the _encode function that do nothing.
class _unicode(object):
pass
_encoding = 'ascii'
def _encode(text, encoding='ascii'):
return text
else:
import locale
_encoding = locale.getpreferredencoding()
def _encode(text, encoding=None):
if isinstance(text, unicode):
return text.encode(encoding or _encoding, 'xmlcharrefreplace')
else:
return text
def _binstr(obj):
# Ensure that we have an encoded (binary) string representation of obj,
# even if it is a unicode string.
return obj.encode(_encoding) if isinstance(obj, _unicode) else str(obj)
# ----------------------------------------------------- module manipulation
def ispackage(path):
@ -424,12 +453,13 @@ class HTMLDoc(Doc):
def page(self, title, contents):
"""Format an HTML page."""
return '''
return _encode('''
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<html><head><title>Python: %s</title>
<meta charset="utf-8">
</head><body bgcolor="#f0f0f8">
%s
</body></html>''' % (title, contents)
</body></html>''' % (title, contents), 'ascii')
def heading(self, title, fgcol, bgcol, extras=''):
"""Format a page heading."""
@ -606,12 +636,12 @@ class HTMLDoc(Doc):
filelink = '(built-in)'
info = []
if hasattr(object, '__version__'):
version = str(object.__version__)
version = _binstr(object.__version__)
if version[:11] == '$' + 'Revision: ' and version[-1:] == '$':
version = strip(version[11:-1])
info.append('version %s' % self.escape(version))
if hasattr(object, '__date__'):
info.append(self.escape(str(object.__date__)))
info.append(self.escape(_binstr(object.__date__)))
if info:
head = head + ' (%s)' % join(info, ', ')
docloc = self.getdocloc(object)
@ -694,11 +724,11 @@ class HTMLDoc(Doc):
result = result + self.bigsection(
'Data', '#ffffff', '#55aa55', join(contents, '<br>\n'))
if hasattr(object, '__author__'):
contents = self.markup(str(object.__author__), self.preformat)
contents = self.markup(_binstr(object.__author__), self.preformat)
result = result + self.bigsection(
'Author', '#ffffff', '#7799ee', contents)
if hasattr(object, '__credits__'):
contents = self.markup(str(object.__credits__), self.preformat)
contents = self.markup(_binstr(object.__credits__), self.preformat)
result = result + self.bigsection(
'Credits', '#ffffff', '#7799ee', contents)
@ -1116,16 +1146,16 @@ class TextDoc(Doc):
result = result + self.section('DATA', join(contents, '\n'))
if hasattr(object, '__version__'):
version = str(object.__version__)
version = _binstr(object.__version__)
if version[:11] == '$' + 'Revision: ' and version[-1:] == '$':
version = strip(version[11:-1])
result = result + self.section('VERSION', version)
if hasattr(object, '__date__'):
result = result + self.section('DATE', str(object.__date__))
result = result + self.section('DATE', _binstr(object.__date__))
if hasattr(object, '__author__'):
result = result + self.section('AUTHOR', str(object.__author__))
result = result + self.section('AUTHOR', _binstr(object.__author__))
if hasattr(object, '__credits__'):
result = result + self.section('CREDITS', str(object.__credits__))
result = result + self.section('CREDITS', _binstr(object.__credits__))
return result
def docclass(self, object, name=None, mod=None, *ignored):
@ -1375,7 +1405,7 @@ def pipepager(text, cmd):
"""Page through text by feeding it to another program."""
pipe = os.popen(cmd, 'w')
try:
pipe.write(text)
pipe.write(_encode(text))
pipe.close()
except IOError:
pass # Ignore broken pipes caused by quitting the pager program.
@ -1385,7 +1415,7 @@ def tempfilepager(text, cmd):
import tempfile
filename = tempfile.mktemp()
file = open(filename, 'w')
file.write(text)
file.write(_encode(text))
file.close()
try:
os.system(cmd + ' "' + filename + '"')
@ -1394,7 +1424,7 @@ def tempfilepager(text, cmd):
def ttypager(text):
"""Page through text on a text terminal."""
lines = split(plain(text), '\n')
lines = plain(_encode(plain(text), getattr(sys.stdout, 'encoding', _encoding))).split('\n')
try:
import tty
fd = sys.stdin.fileno()
@ -1432,7 +1462,7 @@ def ttypager(text):
def plainpager(text):
"""Simply print unformatted text. This is the ultimate fallback."""
sys.stdout.write(plain(text))
sys.stdout.write(_encode(plain(text), getattr(sys.stdout, 'encoding', _encoding)))
def describe(thing):
"""Produce a short description of the given thing."""

View File

@ -10,6 +10,7 @@ import keyword
import pkgutil
import unittest
import xml.etree
import types
import test.test_support
from collections import namedtuple
from test.script_helper import assert_python_ok
@ -428,6 +429,95 @@ class TestDescriptions(unittest.TestCase):
self.assertIn('_asdict', helptext)
@unittest.skipUnless(test.test_support.have_unicode,
"test requires unicode support")
class TestUnicode(unittest.TestCase):
def setUp(self):
# Better not to use unicode escapes in literals, lest the
# parser choke on it if Python has been built without
# unicode support.
self.Q = types.ModuleType(
'Q', 'Rational numbers: \xe2\x84\x9a'.decode('utf8'))
self.Q.__version__ = '\xe2\x84\x9a'.decode('utf8')
self.Q.__date__ = '\xe2\x84\x9a'.decode('utf8')
self.Q.__author__ = '\xe2\x84\x9a'.decode('utf8')
self.Q.__credits__ = '\xe2\x84\x9a'.decode('utf8')
self.assertIsInstance(self.Q.__doc__, unicode)
def test_render_doc(self):
# render_doc is robust against unicode in docstrings
doc = pydoc.render_doc(self.Q)
self.assertIsInstance(doc, str)
def test_encode(self):
# _encode is robust against characters out the specified encoding
self.assertEqual(pydoc._encode(self.Q.__doc__, 'ascii'), 'Rational numbers: &#8474;')
def test_pipepager(self):
# pipepager does not choke on unicode
doc = pydoc.render_doc(self.Q)
saved, os.popen = os.popen, open
try:
with test.test_support.temp_cwd():
pydoc.pipepager(doc, 'pipe')
self.assertEqual(open('pipe').read(), pydoc._encode(doc))
finally:
os.popen = saved
def test_tempfilepager(self):
# tempfilepager does not choke on unicode
doc = pydoc.render_doc(self.Q)
output = {}
def mock_system(cmd):
import ast
output['content'] = open(ast.literal_eval(cmd.strip())).read()
saved, os.system = os.system, mock_system
try:
pydoc.tempfilepager(doc, '')
self.assertEqual(output['content'], pydoc._encode(doc))
finally:
os.system = saved
def test_plainpager(self):
# plainpager does not choke on unicode
doc = pydoc.render_doc(self.Q)
# Note: captured_stdout is too permissive when it comes to
# unicode, and using it here would make the test always
# pass.
with test.test_support.temp_cwd():
with open('output', 'w') as f:
saved, sys.stdout = sys.stdout, f
try:
pydoc.plainpager(doc)
finally:
sys.stdout = saved
self.assertIn('Rational numbers:', open('output').read())
def test_ttypager(self):
# ttypager does not choke on unicode
doc = pydoc.render_doc(self.Q)
# Test ttypager
with test.test_support.temp_cwd(), test.test_support.captured_stdin():
with open('output', 'w') as f:
saved, sys.stdout = sys.stdout, f
try:
pydoc.ttypager(doc)
finally:
sys.stdout = saved
self.assertIn('Rational numbers:', open('output').read())
def test_htmlpage(self):
# html.page does not choke on unicode
with test.test_support.temp_cwd():
with captured_stdout() as output:
pydoc.writedoc(self.Q)
self.assertEqual(output.getvalue(), 'wrote Q.html\n')
class TestHelper(unittest.TestCase):
def test_keywords(self):
self.assertEqual(sorted(pydoc.Helper.keywords),
@ -456,6 +546,7 @@ def test_main():
test.test_support.run_unittest(PydocDocTest,
PydocImportTest,
TestDescriptions,
TestUnicode,
TestHelper)
finally:
reap_children()

View File

@ -30,6 +30,8 @@ Core and Builtins
Library
-------
- Issue #1065986: pydoc can now handle unicode strings.
- Issue #16039: CVE-2013-1752: Change use of readline in imaplib module to
limit line length. Patch by Emil Lind.