Lib/test/test_csv.py

   1 # -*- coding: iso-8859-1 -*-
   2 # Copyright (C) 2001,2002 Python Software Foundation
   3 # csv package unit tests
   4
   5 import sys
   6 import unittest
   7 from StringIO import StringIO
   8 import csv
   9 import gc
  10 from test import test_support
  11
  12 class Test_Csv(unittest.TestCase):
  13     """
  14     Test the underlying C csv parser in ways that are not appropriate
  15     from the high level interface. Further tests of this nature are done
  16     in TestDialectRegistry.
  17     """
  18     def test_reader_arg_valid(self):
  19         self.assertRaises(TypeError, csv.reader)
  20         self.assertRaises(TypeError, csv.reader, None)
  21         self.assertRaises(AttributeError, csv.reader, [], bad_attr = 0)
  22         self.assertRaises(csv.Error, csv.reader, [], 'foo')
  23         class BadClass:
  24             def __init__(self):
  25                 raise IOError
  26         self.assertRaises(IOError, csv.reader, [], BadClass)
  27         self.assertRaises(TypeError, csv.reader, [], None)
  28         class BadDialect:
  29             bad_attr = 0
  30         self.assertRaises(AttributeError, csv.reader, [], BadDialect)
  31
  32     def test_writer_arg_valid(self):
  33         self.assertRaises(TypeError, csv.writer)
  34         self.assertRaises(TypeError, csv.writer, None)
  35         self.assertRaises(AttributeError, csv.writer, StringIO(), bad_attr = 0)
  36
  37     def _test_attrs(self, obj):
  38         self.assertEqual(obj.dialect.delimiter, ',')
  39         obj.dialect.delimiter = '\t'
  40         self.assertEqual(obj.dialect.delimiter, '\t')
  41         self.assertRaises(TypeError, delattr, obj.dialect, 'delimiter')
  42         self.assertRaises(TypeError, setattr, obj.dialect,
  43                           'lineterminator', None)
  44         obj.dialect.escapechar = None
  45         self.assertEqual(obj.dialect.escapechar, None)
  46         self.assertRaises(TypeError, delattr, obj.dialect, 'quoting')
  47         self.assertRaises(TypeError, setattr, obj.dialect, 'quoting', None)
  48         obj.dialect.quoting = csv.QUOTE_MINIMAL
  49         self.assertEqual(obj.dialect.quoting, csv.QUOTE_MINIMAL)
  50
  51     def test_reader_attrs(self):
  52         self._test_attrs(csv.reader([]))
  53
  54     def test_writer_attrs(self):
  55         self._test_attrs(csv.writer(StringIO()))
  56
  57     def _write_test(self, fields, expect, **kwargs):
  58         fileobj = StringIO()
  59         writer = csv.writer(fileobj, **kwargs)
  60         writer.writerow(fields)
  61         self.assertEqual(fileobj.getvalue(),
  62                          expect + writer.dialect.lineterminator)
  63
  64     def test_write_arg_valid(self):
  65         self.assertRaises(csv.Error, self._write_test, None, '')
  66         self._write_test((), '')
  67         self._write_test([None], '""')
  68         self.assertRaises(csv.Error, self._write_test,
  69                           [None], None, quoting = csv.QUOTE_NONE)
  70         # Check that exceptions are passed up the chain
  71         class BadList:
  72             def __len__(self):
  73                 return 10;
  74             def __getitem__(self, i):
  75                 if i > 2:
  76                     raise IOError
  77         self.assertRaises(IOError, self._write_test, BadList(), '')
  78         class BadItem:
  79             def __str__(self):
  80                 raise IOError
  81         self.assertRaises(IOError, self._write_test, [BadItem()], '')
  82
  83     def test_write_bigfield(self):
  84         # This exercises the buffer realloc functionality
  85         bigstring = 'X' * 50000
  86         self._write_test([bigstring,bigstring], '%s,%s' % \
  87                          (bigstring, bigstring))
  88
  89     def test_write_quoting(self):
  90         self._write_test(['a','1','p,q'], 'a,1,"p,q"')
  91         self.assertRaises(csv.Error,
  92                           self._write_test,
  93                           ['a','1','p,q'], 'a,1,"p,q"',
  94                           quoting = csv.QUOTE_NONE)
  95         self._write_test(['a','1','p,q'], 'a,1,"p,q"',
  96                          quoting = csv.QUOTE_MINIMAL)
  97         self._write_test(['a','1','p,q'], '"a",1,"p,q"',
  98                          quoting = csv.QUOTE_NONNUMERIC)
  99         self._write_test(['a','1','p,q'], '"a","1","p,q"',
 100                          quoting = csv.QUOTE_ALL)
 101
 102     def test_write_escape(self):
 103         self._write_test(['a','1','p,q'], 'a,1,"p,q"',
 104                          escapechar='\\')
 105 # FAILED - needs to be fixed [am]:
 106 #        self._write_test(['a','1','p,"q"'], 'a,1,"p,\\"q\\"',
 107 #                         escapechar='\\', doublequote = 0)
 108         self._write_test(['a','1','p,q'], 'a,1,p\\,q',
 109                          escapechar='\\', quoting = csv.QUOTE_NONE)
 110
 111     def test_writerows(self):
 112         class BrokenFile:
 113             def write(self, buf):
 114                 raise IOError
 115         writer = csv.writer(BrokenFile())
 116         self.assertRaises(IOError, writer.writerows, [['a']])
 117         fileobj = StringIO()
 118         writer = csv.writer(fileobj)
 119         self.assertRaises(TypeError, writer.writerows, None)
 120         writer.writerows([['a','b'],['c','d']])
 121         self.assertEqual(fileobj.getvalue(), "a,b\r\nc,d\r\n")
 122
 123     def _read_test(self, input, expect, **kwargs):
 124         reader = csv.reader(input, **kwargs)
 125         result = list(reader)
 126         self.assertEqual(result, expect)
 127
 128     def test_read_oddinputs(self):
 129         self._read_test([], [])
 130         self._read_test([''], [[]])
 131         self.assertRaises(csv.Error, self._read_test,
 132                           ['"ab"c'], None, strict = 1)
 133         # cannot handle null bytes for the moment
 134         self.assertRaises(csv.Error, self._read_test,
 135                           ['ab\0c'], None, strict = 1)
 136         self._read_test(['"ab"c'], [['abc']], doublequote = 0)
 137
 138     def test_read_eol(self):
 139         self._read_test(['a,b'], [['a','b']])
 140         self._read_test(['a,b\n'], [['a','b']])
 141         self._read_test(['a,b\r\n'], [['a','b']])
 142         self._read_test(['a,b\r'], [['a','b']])
 143         self.assertRaises(csv.Error, self._read_test, ['a,b\rc,d'], [])
 144         self.assertRaises(csv.Error, self._read_test, ['a,b\nc,d'], [])
 145         self.assertRaises(csv.Error, self._read_test, ['a,b\r\nc,d'], [])
 146
 147     def test_read_escape(self):
 148         self._read_test(['a,\\b,c'], [['a', '\\b', 'c']], escapechar='\\')
 149         self._read_test(['a,b\\,c'], [['a', 'b,c']], escapechar='\\')
 150         self._read_test(['a,"b\\,c"'], [['a', 'b,c']], escapechar='\\')
 151         self._read_test(['a,"b,\\c"'], [['a', 'b,\\c']], escapechar='\\')
 152         self._read_test(['a,"b,c\\""'], [['a', 'b,c"']], escapechar='\\')
 153         self._read_test(['a,"b,c"\\'], [['a', 'b,c\\']], escapechar='\\')
 154
 155     def test_read_bigfield(self):
 156         # This exercises the buffer realloc functionality
 157         bigstring = 'X' * 50000
 158         bigline = '%s,%s' % (bigstring, bigstring)
 159         self._read_test([bigline], [[bigstring, bigstring]])
 160
 161 class TestDialectRegistry(unittest.TestCase):
 162     def test_registry_badargs(self):
 163         self.assertRaises(TypeError, csv.list_dialects, None)
 164         self.assertRaises(TypeError, csv.get_dialect)
 165         self.assertRaises(csv.Error, csv.get_dialect, None)
 166         self.assertRaises(csv.Error, csv.get_dialect, "nonesuch")
 167         self.assertRaises(TypeError, csv.unregister_dialect)
 168         self.assertRaises(csv.Error, csv.unregister_dialect, None)
 169         self.assertRaises(csv.Error, csv.unregister_dialect, "nonesuch")
 170         self.assertRaises(TypeError, csv.register_dialect, None)
 171         self.assertRaises(TypeError, csv.register_dialect, None, None)
 172         self.assertRaises(TypeError, csv.register_dialect, "nonesuch", None)
 173         class bogus:
 174             def __init__(self):
 175                 raise KeyError
 176         self.assertRaises(KeyError, csv.register_dialect, "nonesuch", bogus)
 177
 178     def test_registry(self):
 179         class myexceltsv(csv.excel):
 180             delimiter = "\t"
 181         name = "myexceltsv"
 182         expected_dialects = csv.list_dialects() + [name]
 183         expected_dialects.sort()
 184         csv.register_dialect(name, myexceltsv)
 185         try:
 186             self.failUnless(isinstance(csv.get_dialect(name), myexceltsv))
 187             got_dialects = csv.list_dialects()
 188             got_dialects.sort()
 189             self.assertEqual(expected_dialects, got_dialects)
 190         finally:
 191             csv.unregister_dialect(name)
 192
 193     def test_incomplete_dialect(self):
 194         class myexceltsv(csv.Dialect):
 195             delimiter = "\t"
 196         self.assertRaises(csv.Error, myexceltsv)
 197
 198     def test_space_dialect(self):
 199         class space(csv.excel):
 200             delimiter = " "
 201             quoting = csv.QUOTE_NONE
 202             escapechar = "\\"
 203
 204         s = StringIO("abc def\nc1ccccc1 benzene\n")
 205         rdr = csv.reader(s, dialect=space())
 206         self.assertEqual(rdr.next(), ["abc", "def"])
 207         self.assertEqual(rdr.next(), ["c1ccccc1", "benzene"])
 208
 209     def test_dialect_apply(self):
 210         class testA(csv.excel):
 211             delimiter = "\t"
 212         class testB(csv.excel):
 213             delimiter = ":"
 214         class testC(csv.excel):
 215             delimiter = "|"
 216
 217         csv.register_dialect('testC', testC)
 218         try:
 219             fileobj = StringIO()
 220             writer = csv.writer(fileobj)
 221             writer.writerow([1,2,3])
 222             self.assertEqual(fileobj.getvalue(), "1,2,3\r\n")
 223
 224             fileobj = StringIO()
 225             writer = csv.writer(fileobj, testA)
 226             writer.writerow([1,2,3])
 227             self.assertEqual(fileobj.getvalue(), "1\t2\t3\r\n")
 228
 229             fileobj = StringIO()
 230             writer = csv.writer(fileobj, dialect=testB())
 231             writer.writerow([1,2,3])
 232             self.assertEqual(fileobj.getvalue(), "1:2:3\r\n")
 233
 234             fileobj = StringIO()
 235             writer = csv.writer(fileobj, dialect='testC')
 236             writer.writerow([1,2,3])
 237             self.assertEqual(fileobj.getvalue(), "1|2|3\r\n")
 238
 239             fileobj = StringIO()
 240             writer = csv.writer(fileobj, dialect=testA, delimiter=';')
 241             writer.writerow([1,2,3])
 242             self.assertEqual(fileobj.getvalue(), "1;2;3\r\n")
 243         finally:
 244             csv.unregister_dialect('testC')
 245
 246     def test_bad_dialect(self):
 247         # Unknown parameter
 248         self.assertRaises(AttributeError, csv.reader, [], bad_attr = 0)
 249         # Bad values
 250         self.assertRaises(TypeError, csv.reader, [], delimiter = None)
 251         self.assertRaises(TypeError, csv.reader, [], quoting = -1)
 252         self.assertRaises(TypeError, csv.reader, [], quoting = 100)
 253
 254 class TestCsvBase(unittest.TestCase):
 255     def readerAssertEqual(self, input, expected_result):
 256         reader = csv.reader(StringIO(input), dialect = self.dialect)
 257         fields = list(reader)
 258         self.assertEqual(fields, expected_result)
 259
 260     def writerAssertEqual(self, input, expected_result):
 261         fileobj = StringIO()
 262         writer = csv.writer(fileobj, dialect = self.dialect)
 263         writer.writerows(input)
 264         self.assertEqual(fileobj.getvalue(), expected_result)
 265
 266 class TestDialectExcel(TestCsvBase):
 267     dialect = 'excel'
 268
 269     def test_single(self):
 270         self.readerAssertEqual('abc', [['abc']])
 271
 272     def test_simple(self):
 273         self.readerAssertEqual('1,2,3,4,5', [['1','2','3','4','5']])
 274
 275     def test_blankline(self):
 276         self.readerAssertEqual('', [])
 277
 278     def test_empty_fields(self):
 279         self.readerAssertEqual(',', [['', '']])
 280
 281     def test_singlequoted(self):
 282         self.readerAssertEqual('""', [['']])
 283
 284     def test_singlequoted_left_empty(self):
 285         self.readerAssertEqual('"",', [['','']])
 286
 287     def test_singlequoted_right_empty(self):
 288         self.readerAssertEqual(',""', [['','']])
 289
 290     def test_single_quoted_quote(self):
 291         self.readerAssertEqual('""""', [['"']])
 292
 293     def test_quoted_quotes(self):
 294         self.readerAssertEqual('""""""', [['""']])
 295
 296     def test_inline_quote(self):
 297         self.readerAssertEqual('a""b', [['a""b']])
 298
 299     def test_inline_quotes(self):
 300         self.readerAssertEqual('a"b"c', [['a"b"c']])
 301
 302     def test_quotes_and_more(self):
 303         self.readerAssertEqual('"a"b', [['ab']])
 304
 305     def test_lone_quote(self):
 306         self.readerAssertEqual('a"b', [['a"b']])
 307
 308     def test_quote_and_quote(self):
 309         self.readerAssertEqual('"a" "b"', [['a "b"']])
 310
 311     def test_space_and_quote(self):
 312         self.readerAssertEqual(' "a"', [[' "a"']])
 313
 314     def test_quoted(self):
 315         self.readerAssertEqual('1,2,3,"I think, therefore I am",5,6',
 316                                [['1', '2', '3',
 317                                  'I think, therefore I am',
 318                                  '5', '6']])
 319
 320     def test_quoted_quote(self):
 321         self.readerAssertEqual('1,2,3,"""I see,"" said the blind man","as he picked up his hammer and saw"',
 322                                [['1', '2', '3',
 323                                  '"I see," said the blind man',
 324                                  'as he picked up his hammer and saw']])
 325
 326     def test_quoted_nl(self):
 327         input = '''\
 328 1,2,3,"""I see,""
 329 said the blind man","as he picked up his
 330 hammer and saw"
 331 9,8,7,6'''
 332         self.readerAssertEqual(input,
 333                                [['1', '2', '3',
 334                                    '"I see,"\nsaid the blind man',
 335                                    'as he picked up his\nhammer and saw'],
 336                                 ['9','8','7','6']])
 337
 338     def test_dubious_quote(self):
 339         self.readerAssertEqual('12,12,1",', [['12', '12', '1"', '']])
 340
 341     def test_null(self):
 342         self.writerAssertEqual([], '')
 343
 344     def test_single(self):
 345         self.writerAssertEqual([['abc']], 'abc\r\n')
 346
 347     def test_simple(self):
 348         self.writerAssertEqual([[1, 2, 'abc', 3, 4]], '1,2,abc,3,4\r\n')
 349
 350     def test_quotes(self):
 351         self.writerAssertEqual([[1, 2, 'a"bc"', 3, 4]], '1,2,"a""bc""",3,4\r\n')
 352
 353     def test_quote_fieldsep(self):
 354         self.writerAssertEqual([['abc,def']], '"abc,def"\r\n')
 355
 356     def test_newlines(self):
 357         self.writerAssertEqual([[1, 2, 'a\nbc', 3, 4]], '1,2,"a\nbc",3,4\r\n')
 358
 359 class EscapedExcel(csv.excel):
 360     quoting = csv.QUOTE_NONE
 361     escapechar = '\\'
 362
 363 class TestEscapedExcel(TestCsvBase):
 364     dialect = EscapedExcel()
 365
 366     def test_escape_fieldsep(self):
 367         self.writerAssertEqual([['abc,def']], 'abc\\,def\r\n')
 368
 369     def test_read_escape_fieldsep(self):
 370         self.readerAssertEqual('abc\\,def\r\n', [['abc,def']])
 371
 372 class QuotedEscapedExcel(csv.excel):
 373     quoting = csv.QUOTE_NONNUMERIC
 374     escapechar = '\\'
 375
 376 class TestQuotedEscapedExcel(TestCsvBase):
 377     dialect = QuotedEscapedExcel()
 378
 379     def test_write_escape_fieldsep(self):
 380         self.writerAssertEqual([['abc,def']], '"abc,def"\r\n')
 381
 382     def test_read_escape_fieldsep(self):
 383         self.readerAssertEqual('"abc\\,def"\r\n', [['abc,def']])
 384
 385 # Disabled, pending support in csv.utils module
 386 class TestDictFields(unittest.TestCase):
 387     ### "long" means the row is longer than the number of fieldnames
 388     ### "short" means there are fewer elements in the row than fieldnames
 389     def test_write_simple_dict(self):
 390         fileobj = StringIO()
 391         writer = csv.DictWriter(fileobj, fieldnames = ["f1", "f2", "f3"])
 392         writer.writerow({"f1": 10, "f3": "abc"})
 393         self.assertEqual(fileobj.getvalue(), "10,,abc\r\n")
 394
 395     def test_write_no_fields(self):
 396         fileobj = StringIO()
 397         self.assertRaises(TypeError, csv.DictWriter, fileobj)
 398
 399     def test_read_dict_fields(self):
 400         reader = csv.DictReader(StringIO("1,2,abc\r\n"),
 401                                 fieldnames=["f1", "f2", "f3"])
 402         self.assertEqual(reader.next(), {"f1": '1', "f2": '2', "f3": 'abc'})
 403
 404     def test_read_long(self):
 405         reader = csv.DictReader(StringIO("1,2,abc,4,5,6\r\n"),
 406                                 fieldnames=["f1", "f2"])
 407         self.assertEqual(reader.next(), {"f1": '1', "f2": '2',
 408                                          None: ["abc", "4", "5", "6"]})
 409
 410     def test_read_long_with_rest(self):
 411         reader = csv.DictReader(StringIO("1,2,abc,4,5,6\r\n"),
 412                                 fieldnames=["f1", "f2"], restkey="_rest")
 413         self.assertEqual(reader.next(), {"f1": '1', "f2": '2',
 414                                          "_rest": ["abc", "4", "5", "6"]})
 415
 416     def test_read_short(self):
 417         reader = csv.DictReader(["1,2,abc,4,5,6\r\n","1,2,abc\r\n"],
 418                                 fieldnames="1 2 3 4 5 6".split(),
 419                                 restval="DEFAULT")
 420         self.assertEqual(reader.next(), {"1": '1', "2": '2', "3": 'abc',
 421                                          "4": '4', "5": '5', "6": '6'})
 422         self.assertEqual(reader.next(), {"1": '1', "2": '2', "3": 'abc',
 423                                          "4": 'DEFAULT', "5": 'DEFAULT',
 424                                          "6": 'DEFAULT'})
 425
 426     def test_read_multi(self):
 427         sample = [
 428             '2147483648,43.0e12,17,abc,def\r\n',
 429             '147483648,43.0e2,17,abc,def\r\n',
 430             '47483648,43.0,170,abc,def\r\n'
 431             ]
 432
 433         reader = csv.DictReader(sample,
 434                                 fieldnames="i1 float i2 s1 s2".split())
 435         self.assertEqual(reader.next(), {"i1": '2147483648',
 436                                          "float": '43.0e12',
 437                                          "i2": '17',
 438                                          "s1": 'abc',
 439                                          "s2": 'def'})
 440
 441     def test_read_with_blanks(self):
 442         reader = csv.DictReader(["1,2,abc,4,5,6\r\n","\r\n",
 443                                  "1,2,abc,4,5,6\r\n"],
 444                                 fieldnames="1 2 3 4 5 6".split())
 445         self.assertEqual(reader.next(), {"1": '1', "2": '2', "3": 'abc',
 446                                          "4": '4', "5": '5', "6": '6'})
 447         self.assertEqual(reader.next(), {"1": '1', "2": '2', "3": 'abc',
 448                                          "4": '4', "5": '5', "6": '6'})
 449
 450 class TestArrayWrites(unittest.TestCase):
 451     def test_int_write(self):
 452         import array
 453         contents = [(20-i) for i in range(20)]
 454         a = array.array('i', contents)
 455         fileobj = StringIO()
 456         writer = csv.writer(fileobj, dialect="excel")
 457         writer.writerow(a)
 458         expected = ",".join([str(i) for i in a])+"\r\n"
 459         self.assertEqual(fileobj.getvalue(), expected)
 460
 461     def test_double_write(self):
 462         import array
 463         contents = [(20-i)*0.1 for i in range(20)]
 464         a = array.array('d', contents)
 465         fileobj = StringIO()
 466         writer = csv.writer(fileobj, dialect="excel")
 467         writer.writerow(a)
 468         expected = ",".join([str(i) for i in a])+"\r\n"
 469         self.assertEqual(fileobj.getvalue(), expected)
 470
 471     def test_float_write(self):
 472         import array
 473         contents = [(20-i)*0.1 for i in range(20)]
 474         a = array.array('f', contents)
 475         fileobj = StringIO()
 476         writer = csv.writer(fileobj, dialect="excel")
 477         writer.writerow(a)
 478         expected = ",".join([str(i) for i in a])+"\r\n"
 479         self.assertEqual(fileobj.getvalue(), expected)
 480
 481     def test_char_write(self):
 482         import array, string
 483         a = array.array('c', string.letters)
 484         fileobj = StringIO()
 485         writer = csv.writer(fileobj, dialect="excel")
 486         writer.writerow(a)
 487         expected = ",".join(a)+"\r\n"
 488         self.assertEqual(fileobj.getvalue(), expected)
 489
 490 class TestDialectValidity(unittest.TestCase):
 491     def test_quoting(self):
 492         class mydialect(csv.Dialect):
 493             delimiter = ";"
 494             escapechar = '\\'
 495             doublequote = False
 496             skipinitialspace = True
 497             lineterminator = '\r\n'
 498             quoting = csv.QUOTE_NONE
 499         d = mydialect()
 500
 501         mydialect.quoting = None
 502         self.assertRaises(csv.Error, mydialect)
 503
 504         mydialect.quoting = csv.QUOTE_NONE
 505         mydialect.escapechar = None
 506         self.assertRaises(csv.Error, mydialect)
 507
 508         mydialect.doublequote = True
 509         mydialect.quoting = csv.QUOTE_ALL
 510         mydialect.quotechar = '"'
 511         d = mydialect()
 512
 513         mydialect.quotechar = "''"
 514         self.assertRaises(csv.Error, mydialect)
 515
 516         mydialect.quotechar = 4
 517         self.assertRaises(csv.Error, mydialect)
 518
 519     def test_delimiter(self):
 520         class mydialect(csv.Dialect):
 521             delimiter = ";"
 522             escapechar = '\\'
 523             doublequote = False
 524             skipinitialspace = True
 525             lineterminator = '\r\n'
 526             quoting = csv.QUOTE_NONE
 527         d = mydialect()
 528
 529         mydialect.delimiter = ":::"
 530         self.assertRaises(csv.Error, mydialect)
 531
 532         mydialect.delimiter = 4
 533         self.assertRaises(csv.Error, mydialect)
 534
 535     def test_lineterminator(self):
 536         class mydialect(csv.Dialect):
 537             delimiter = ";"
 538             escapechar = '\\'
 539             doublequote = False
 540             skipinitialspace = True
 541             lineterminator = '\r\n'
 542             quoting = csv.QUOTE_NONE
 543         d = mydialect()
 544
 545         mydialect.lineterminator = ":::"
 546         d = mydialect()
 547
 548         mydialect.lineterminator = 4
 549         self.assertRaises(csv.Error, mydialect)
 550
 551
 552 class TestSniffer(unittest.TestCase):
 553     sample1 = """\
 554 Harry's, Arlington Heights, IL, 2/1/03, Kimi Hayes
 555 Shark City, Glendale Heights, IL, 12/28/02, Prezence
 556 Tommy's Place, Blue Island, IL, 12/28/02, Blue Sunday/White Crow
 557 Stonecutters Seafood and Chop House, Lemont, IL, 12/19/02, Week Back
 558 """
 559     sample2 = """\
 560 'Harry''s':'Arlington Heights':'IL':'2/1/03':'Kimi Hayes'
 561 'Shark City':'Glendale Heights':'IL':'12/28/02':'Prezence'
 562 'Tommy''s Place':'Blue Island':'IL':'12/28/02':'Blue Sunday/White Crow'
 563 'Stonecutters Seafood and Chop House':'Lemont':'IL':'12/19/02':'Week Back'
 564 """
 565
 566     header = '''\
 567 "venue","city","state","date","performers"
 568 '''
 569     sample3 = '''\
 570 05/05/03?05/05/03?05/05/03?05/05/03?05/05/03?05/05/03
 571 05/05/03?05/05/03?05/05/03?05/05/03?05/05/03?05/05/03
 572 05/05/03?05/05/03?05/05/03?05/05/03?05/05/03?05/05/03
 573 '''
 574
 575     sample4 = '''\
 576 2147483648;43.0e12;17;abc;def
 577 147483648;43.0e2;17;abc;def
 578 47483648;43.0;170;abc;def
 579 '''
 580
 581     def test_has_header(self):
 582         sniffer = csv.Sniffer()
 583         self.assertEqual(sniffer.has_header(self.sample1), False)
 584         self.assertEqual(sniffer.has_header(self.header+self.sample1), True)
 585
 586     def test_sniff(self):
 587         sniffer = csv.Sniffer()
 588         dialect = sniffer.sniff(self.sample1)
 589         self.assertEqual(dialect.delimiter, ",")
 590         self.assertEqual(dialect.quotechar, '"')
 591         self.assertEqual(dialect.skipinitialspace, True)
 592
 593         dialect = sniffer.sniff(self.sample2)
 594         self.assertEqual(dialect.delimiter, ":")
 595         self.assertEqual(dialect.quotechar, "'")
 596         self.assertEqual(dialect.skipinitialspace, False)
 597
 598     def test_delimiters(self):
 599         sniffer = csv.Sniffer()
 600         dialect = sniffer.sniff(self.sample3)
 601         self.assertEqual(dialect.delimiter, "0")
 602         dialect = sniffer.sniff(self.sample3, delimiters="?,")
 603         self.assertEqual(dialect.delimiter, "?")
 604         dialect = sniffer.sniff(self.sample3, delimiters="/,")
 605         self.assertEqual(dialect.delimiter, "/")
 606         dialect = sniffer.sniff(self.sample4)
 607         self.assertEqual(dialect.delimiter, ";")
 608
 609 if not hasattr(sys, "gettotalrefcount"):
 610     if test_support.verbose: print "*** skipping leakage tests ***"
 611 else:
 612     class NUL:
 613         def write(s, *args):
 614             pass
 615         writelines = write
 616
 617     class TestLeaks(unittest.TestCase):
 618         def test_create_read(self):
 619             delta = 0
 620             lastrc = sys.gettotalrefcount()
 621             for i in xrange(20):
 622                 gc.collect()
 623                 self.assertEqual(gc.garbage, [])
 624                 rc = sys.gettotalrefcount()
 625                 csv.reader(["a,b,c\r\n"])
 626                 csv.reader(["a,b,c\r\n"])
 627                 csv.reader(["a,b,c\r\n"])
 628                 delta = rc-lastrc
 629                 lastrc = rc
 630             # if csv.reader() leaks, last delta should be 3 or more
 631             self.assertEqual(delta < 3, True)
 632
 633         def test_create_write(self):
 634             delta = 0
 635             lastrc = sys.gettotalrefcount()
 636             s = NUL()
 637             for i in xrange(20):
 638                 gc.collect()
 639                 self.assertEqual(gc.garbage, [])
 640                 rc = sys.gettotalrefcount()
 641                 csv.writer(s)
 642                 csv.writer(s)
 643                 csv.writer(s)
 644                 delta = rc-lastrc
 645                 lastrc = rc
 646             # if csv.writer() leaks, last delta should be 3 or more
 647             self.assertEqual(delta < 3, True)
 648
 649         def test_read(self):
 650             delta = 0
 651             rows = ["a,b,c\r\n"]*5
 652             lastrc = sys.gettotalrefcount()
 653             for i in xrange(20):
 654                 gc.collect()
 655                 self.assertEqual(gc.garbage, [])
 656                 rc = sys.gettotalrefcount()
 657                 rdr = csv.reader(rows)
 658                 for row in rdr:
 659                     pass
 660                 delta = rc-lastrc
 661                 lastrc = rc
 662             # if reader leaks during read, delta should be 5 or more
 663             self.assertEqual(delta < 5, True)
 664
 665         def test_write(self):
 666             delta = 0
 667             rows = [[1,2,3]]*5
 668             s = NUL()
 669             lastrc = sys.gettotalrefcount()
 670             for i in xrange(20):
 671                 gc.collect()
 672                 self.assertEqual(gc.garbage, [])
 673                 rc = sys.gettotalrefcount()
 674                 writer = csv.writer(s)
 675                 for row in rows:
 676                     writer.writerow(row)
 677                 delta = rc-lastrc
 678                 lastrc = rc
 679             # if writer leaks during write, last delta should be 5 or more
 680             self.assertEqual(delta < 5, True)
 681
 682 # commented out for now - csv module doesn't yet support Unicode
 683 if 0:
 684     from StringIO import StringIO
 685     import csv
 686
 687     class TestUnicode(unittest.TestCase):
 688         def test_unicode_read(self):
 689             import codecs
 690             f = codecs.EncodedFile(StringIO("Martin von Löwis,"
 691                                             "Marc André Lemburg,"
 692                                             "Guido van Rossum,"
 693                                             "François Pinard\r\n"),
 694                                    data_encoding='iso-8859-1')
 695             reader = csv.reader(f)
 696             self.assertEqual(list(reader), [[u"Martin von Löwis",
 697                                              u"Marc André Lemburg",
 698                                              u"Guido van Rossum",
 699                                              u"François Pinardn"]])
 700
 701 def test_main():
 702     mod = sys.modules[__name__]
 703     test_support.run_unittest(
 704         *[getattr(mod, name) for name in dir(mod) if name.startswith('Test')]
 705     )
 706
 707 if __name__ == '__main__':
 708     test_main()