test/test_io.py

   1 #! /usr/bin/env python
   2
   3 # $Id$
   4 # Author: Lea Wiemann <LeWiemann@gmail.com>
   5 # Copyright: This module has been placed in the public domain.
   6
   7 """
   8 Test module for io.py.
   9 """
  10
  11 import unittest, sys
  12 import DocutilsTestSupport              # must be imported before docutils
  13 from docutils import io
  14 from docutils._compat import b, bytes
  15 from docutils.utils.error_reporting import locale_encoding
  16 from test_error_reporting import BBuf, UBuf
  17
  18 class mock_stdout(UBuf):
  19     encoding = 'utf8'
  20
  21     def __init__(self):
  22         self.buffer = BBuf()
  23         UBuf.__init__(self)
  24
  25 class HelperTests(unittest.TestCase):
  26
  27     def test_check_encoding_true(self):
  28         """Return `True` if lookup returns the same codec"""
  29         self.assertEqual(io.check_encoding(mock_stdout, 'utf8'), True)
  30         self.assertEqual(io.check_encoding(mock_stdout, 'utf-8'), True)
  31         self.assertEqual(io.check_encoding(mock_stdout, 'UTF-8'), True)
  32
  33     def test_check_encoding_false(self):
  34         """Return `False` if lookup returns different codecs"""
  35         self.assertEqual(io.check_encoding(mock_stdout, 'ascii'), False)
  36         self.assertEqual(io.check_encoding(mock_stdout, 'latin-1'), False)
  37
  38     def test_check_encoding_none(self):
  39         """Cases where the comparison fails."""
  40         # stream.encoding is None:
  41         self.assertEqual(io.check_encoding(io.FileInput(), 'ascii'), None)
  42         # stream.encoding does not exist:
  43         self.assertEqual(io.check_encoding(BBuf, 'ascii'), None)
  44         # encoding is None:
  45         self.assertEqual(io.check_encoding(mock_stdout, None), None)
  46         # encoding is invalid
  47         self.assertEqual(io.check_encoding(mock_stdout, 'UTF-9'), None)
  48
  49
  50 class InputTests(unittest.TestCase):
  51
  52     def test_bom(self):
  53         input = io.StringInput(source=b('\xef\xbb\xbf foo \xef\xbb\xbf bar'),
  54                                encoding='utf8')
  55         # Assert BOMs are gone.
  56         self.assertEqual(input.read(), u' foo  bar')
  57         # With unicode input:
  58         input = io.StringInput(source=u'\ufeff foo \ufeff bar')
  59         # Assert BOMs are still there.
  60         self.assertEqual(input.read(), u'\ufeff foo \ufeff bar')
  61
  62     def test_coding_slug(self):
  63         input = io.StringInput(source=b("""\
  64 .. -*- coding: ascii -*-
  65 data
  66 blah
  67 """))
  68         data = input.read()
  69         self.assertEqual(input.successful_encoding, 'ascii')
  70         input = io.StringInput(source=b("""\
  71 #! python
  72 # -*- coding: ascii -*-
  73 print "hello world"
  74 """))
  75         data = input.read()
  76         self.assertEqual(input.successful_encoding, 'ascii')
  77         input = io.StringInput(source=b("""\
  78 #! python
  79 # extraneous comment; prevents coding slug from being read
  80 # -*- coding: ascii -*-
  81 print "hello world"
  82 """))
  83         data = input.read()
  84         self.assertNotEqual(input.successful_encoding, 'ascii')
  85
  86     def test_bom_detection(self):
  87         source = u'\ufeffdata\nblah\n'
  88         input = io.StringInput(source=source.encode('utf-16-be'))
  89         data = input.read()
  90         self.assertEqual(input.successful_encoding, 'utf-16-be')
  91         input = io.StringInput(source=source.encode('utf-16-le'))
  92         data = input.read()
  93         self.assertEqual(input.successful_encoding, 'utf-16-le')
  94         input = io.StringInput(source=source.encode('utf-8'))
  95         data = input.read()
  96         self.assertEqual(input.successful_encoding, 'utf-8')
  97
  98     def test_readlines(self):
  99         input = io.FileInput(source_path='data/include.txt')
 100         data = input.readlines()
 101         self.assertEqual(data, [u'Some include text.\n'])
 102
 103     def test_heuristics_utf8(self):
 104         # if no encoding is given, try decoding with utf8:
 105         input = io.FileInput(source_path='functional/input/cyrillic.txt')
 106         data = input.read()
 107         if sys.version_info < (3,0):
 108             # in Py3k, the locale encoding is used without --input-encoding
 109             # skipping the heuristic
 110             self.assertEqual(input.successful_encoding, 'utf-8')
 111
 112     def test_heuristics_no_utf8(self):
 113         # if no encoding is given and decoding with utf8 fails,
 114         # use either the locale encoding (if specified) or latin1:
 115         input = io.FileInput(source_path='data/latin1.txt')
 116         data = input.read()
 117         self.assertTrue(input.successful_encoding in (locale_encoding,
 118                                                       'latin-1'))
 119         if input.successful_encoding == 'latin-1':
 120             self.assertEqual(data, u'Gr\xfc\xdfe\n')
 121
 122
 123 class OutputTests(unittest.TestCase):
 124
 125     bdata = b('\xfc')
 126     udata = u'\xfc'
 127
 128     def setUp(self):
 129         self.bdrain = BBuf()
 130         """Buffer accepting binary strings (bytes)"""
 131         self.udrain = UBuf()
 132         """Buffer accepting unicode strings"""
 133         self.mock_stdout = mock_stdout()
 134         """Stub of sys.stdout under Python 3"""
 135
 136     def test_write_unicode(self):
 137         fo = io.FileOutput(destination=self.udrain, encoding='unicode',
 138                            autoclose=False)
 139         fo.write(self.udata)
 140         self.assertEqual(self.udrain.getvalue(), self.udata)
 141
 142     def test_write_utf8(self):
 143         if sys.version_info >= (3,0):
 144             fo = io.FileOutput(destination=self.udrain, encoding='utf8',
 145                                autoclose=False)
 146             fo.write(self.udata)
 147             self.assertEqual(self.udrain.getvalue(), self.udata)
 148         else:
 149             fo = io.FileOutput(destination=self.bdrain, encoding='utf8',
 150                                autoclose=False)
 151             fo.write(self.udata)
 152             self.assertEqual(self.bdrain.getvalue(), self.udata.encode('utf8'))
 153
 154     # With destination in binary mode, data must be binary string
 155     # and is written as-is:
 156     def test_write_bytes(self):
 157         fo = io.FileOutput(destination=self.bdrain, encoding='utf8',
 158                            mode='wb', autoclose=False)
 159         fo.write(self.bdata)
 160         self.assertEqual(self.bdrain.getvalue(), self.bdata)
 161
 162     # Test for Python 3 features:
 163     if sys.version_info >= (3,0):
 164         def test_write_bytes_to_stdout(self):
 165             # binary data is written to destination.buffer, if the
 166             # destination is sys.stdout or sys.stdin
 167             backup = sys.stdout
 168             sys.stdout = self.mock_stdout
 169             fo = io.FileOutput(destination=sys.stdout, mode='wb',
 170                                autoclose=False)
 171             fo.write(self.bdata)
 172             self.assertEqual(self.mock_stdout.buffer.getvalue(),
 173                              self.bdata)
 174             sys.stdout = backup
 175
 176         def test_encoding_clash(self):
 177             # Raise error, if given and destination encodings differ
 178             # TODO: try the `write to .buffer` scheme instead?
 179             self.assertRaises(ValueError,
 180                               io.FileOutput, destination=self.mock_stdout,
 181                                encoding='latin1')
 182
 183
 184 if __name__ == '__main__':
 185     unittest.main()