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     def test_read_semi_sep(self):
 451         reader = csv.DictReader(["1;2;abc;4;5;6\r\n"],
 452                                 fieldnames="1 2 3 4 5 6".split(),
 453                                 delimiter=';')
 454         self.assertEqual(reader.next(), {"1": '1', "2": '2', "3": 'abc',
 455                                          "4": '4', "5": '5', "6": '6'})
 456
 457 class TestArrayWrites(unittest.TestCase):
 458     def test_int_write(self):
 459         import array
 460         contents = [(20-i) for i in range(20)]
 461         a = array.array('i', contents)
 462         fileobj = StringIO()
 463         writer = csv.writer(fileobj, dialect="excel")
 464         writer.writerow(a)
 465         expected = ",".join([str(i) for i in a])+"\r\n"
 466         self.assertEqual(fileobj.getvalue(), expected)
 467
 468     def test_double_write(self):
 469         import array
 470         contents = [(20-i)*0.1 for i in range(20)]
 471         a = array.array('d', contents)
 472         fileobj = StringIO()
 473         writer = csv.writer(fileobj, dialect="excel")
 474         writer.writerow(a)
 475         expected = ",".join([str(i) for i in a])+"\r\n"
 476         self.assertEqual(fileobj.getvalue(), expected)
 477
 478     def test_float_write(self):
 479         import array
 480         contents = [(20-i)*0.1 for i in range(20)]
 481         a = array.array('f', contents)
 482         fileobj = StringIO()
 483         writer = csv.writer(fileobj, dialect="excel")
 484         writer.writerow(a)
 485         expected = ",".join([str(i) for i in a])+"\r\n"
 486         self.assertEqual(fileobj.getvalue(), expected)
 487
 488     def test_char_write(self):
 489         import array, string
 490         a = array.array('c', string.letters)
 491         fileobj = StringIO()
 492         writer = csv.writer(fileobj, dialect="excel")
 493         writer.writerow(a)
 494         expected = ",".join(a)+"\r\n"
 495         self.assertEqual(fileobj.getvalue(), expected)
 496
 497 class TestDialectValidity(unittest.TestCase):
 498     def test_quoting(self):
 499         class mydialect(csv.Dialect):
 500             delimiter = ";"
 501             escapechar = '\\'
 502             doublequote = False
 503             skipinitialspace = True
 504             lineterminator = '\r\n'
 505             quoting = csv.QUOTE_NONE
 506         d = mydialect()
 507
 508         mydialect.quoting = None
 509         self.assertRaises(csv.Error, mydialect)
 510
 511         mydialect.quoting = csv.QUOTE_NONE
 512         mydialect.escapechar = None
 513         self.assertRaises(csv.Error, mydialect)
 514
 515         mydialect.doublequote = True
 516         mydialect.quoting = csv.QUOTE_ALL
 517         mydialect.quotechar = '"'
 518         d = mydialect()
 519
 520         mydialect.quotechar = "''"
 521         self.assertRaises(csv.Error, mydialect)
 522
 523         mydialect.quotechar = 4
 524         self.assertRaises(csv.Error, mydialect)
 525
 526     def test_delimiter(self):
 527         class mydialect(csv.Dialect):
 528             delimiter = ";"
 529             escapechar = '\\'
 530             doublequote = False
 531             skipinitialspace = True
 532             lineterminator = '\r\n'
 533             quoting = csv.QUOTE_NONE
 534         d = mydialect()
 535
 536         mydialect.delimiter = ":::"
 537         self.assertRaises(csv.Error, mydialect)
 538
 539         mydialect.delimiter = 4
 540         self.assertRaises(csv.Error, mydialect)
 541
 542     def test_lineterminator(self):
 543         class mydialect(csv.Dialect):
 544             delimiter = ";"
 545             escapechar = '\\'
 546             doublequote = False
 547             skipinitialspace = True
 548             lineterminator = '\r\n'
 549             quoting = csv.QUOTE_NONE
 550         d = mydialect()
 551
 552         mydialect.lineterminator = ":::"
 553         d = mydialect()
 554
 555         mydialect.lineterminator = 4
 556         self.assertRaises(csv.Error, mydialect)
 557
 558
 559 class TestSniffer(unittest.TestCase):
 560     sample1 = """\
 561 Harry's, Arlington Heights, IL, 2/1/03, Kimi Hayes
 562 Shark City, Glendale Heights, IL, 12/28/02, Prezence
 563 Tommy's Place, Blue Island, IL, 12/28/02, Blue Sunday/White Crow
 564 Stonecutters Seafood and Chop House, Lemont, IL, 12/19/02, Week Back
 565 """
 566     sample2 = """\
 567 'Harry''s':'Arlington Heights':'IL':'2/1/03':'Kimi Hayes'
 568 'Shark City':'Glendale Heights':'IL':'12/28/02':'Prezence'
 569 'Tommy''s Place':'Blue Island':'IL':'12/28/02':'Blue Sunday/White Crow'
 570 'Stonecutters Seafood and Chop House':'Lemont':'IL':'12/19/02':'Week Back'
 571 """
 572
 573     header = '''\
 574 "venue","city","state","date","performers"
 575 '''
 576     sample3 = '''\
 577 05/05/03?05/05/03?05/05/03?05/05/03?05/05/03?05/05/03
 578 05/05/03?05/05/03?05/05/03?05/05/03?05/05/03?05/05/03
 579 05/05/03?05/05/03?05/05/03?05/05/03?05/05/03?05/05/03
 580 '''
 581
 582     sample4 = '''\
 583 2147483648;43.0e12;17;abc;def
 584 147483648;43.0e2;17;abc;def
 585 47483648;43.0;170;abc;def
 586 '''
 587
 588     def test_has_header(self):
 589         sniffer = csv.Sniffer()
 590         self.assertEqual(sniffer.has_header(self.sample1), False)
 591         self.assertEqual(sniffer.has_header(self.header+self.sample1), True)
 592
 593     def test_sniff(self):
 594         sniffer = csv.Sniffer()
 595         dialect = sniffer.sniff(self.sample1)
 596         self.assertEqual(dialect.delimiter, ",")
 597         self.assertEqual(dialect.quotechar, '"')
 598         self.assertEqual(dialect.skipinitialspace, True)
 599
 600         dialect = sniffer.sniff(self.sample2)
 601         self.assertEqual(dialect.delimiter, ":")
 602         self.assertEqual(dialect.quotechar, "'")
 603         self.assertEqual(dialect.skipinitialspace, False)
 604
 605     def test_delimiters(self):
 606         sniffer = csv.Sniffer()
 607         dialect = sniffer.sniff(self.sample3)
 608         self.assertEqual(dialect.delimiter, "0")
 609         dialect = sniffer.sniff(self.sample3, delimiters="?,")
 610         self.assertEqual(dialect.delimiter, "?")
 611         dialect = sniffer.sniff(self.sample3, delimiters="/,")
 612         self.assertEqual(dialect.delimiter, "/")
 613         dialect = sniffer.sniff(self.sample4)
 614         self.assertEqual(dialect.delimiter, ";")
 615
 616 if not hasattr(sys, "gettotalrefcount"):
 617     if test_support.verbose: print "*** skipping leakage tests ***"
 618 else:
 619     class NUL:
 620         def write(s, *args):
 621             pass
 622         writelines = write
 623
 624     class TestLeaks(unittest.TestCase):
 625         def test_create_read(self):
 626             delta = 0
 627             lastrc = sys.gettotalrefcount()
 628             for i in xrange(20):
 629                 gc.collect()
 630                 self.assertEqual(gc.garbage, [])
 631                 rc = sys.gettotalrefcount()
 632                 csv.reader(["a,b,c\r\n"])
 633                 csv.reader(["a,b,c\r\n"])
 634                 csv.reader(["a,b,c\r\n"])
 635                 delta = rc-lastrc
 636                 lastrc = rc
 637             # if csv.reader() leaks, last delta should be 3 or more
 638             self.assertEqual(delta < 3, True)
 639
 640         def test_create_write(self):
 641             delta = 0
 642             lastrc = sys.gettotalrefcount()
 643             s = NUL()
 644             for i in xrange(20):
 645                 gc.collect()
 646                 self.assertEqual(gc.garbage, [])
 647                 rc = sys.gettotalrefcount()
 648                 csv.writer(s)
 649                 csv.writer(s)
 650                 csv.writer(s)
 651                 delta = rc-lastrc
 652                 lastrc = rc
 653             # if csv.writer() leaks, last delta should be 3 or more
 654             self.assertEqual(delta < 3, True)
 655
 656         def test_read(self):
 657             delta = 0
 658             rows = ["a,b,c\r\n"]*5
 659             lastrc = sys.gettotalrefcount()
 660             for i in xrange(20):
 661                 gc.collect()
 662                 self.assertEqual(gc.garbage, [])
 663                 rc = sys.gettotalrefcount()
 664                 rdr = csv.reader(rows)
 665                 for row in rdr:
 666                     pass
 667                 delta = rc-lastrc
 668                 lastrc = rc
 669             # if reader leaks during read, delta should be 5 or more
 670             self.assertEqual(delta < 5, True)
 671
 672         def test_write(self):
 673             delta = 0
 674             rows = [[1,2,3]]*5
 675             s = NUL()
 676             lastrc = sys.gettotalrefcount()
 677             for i in xrange(20):
 678                 gc.collect()
 679                 self.assertEqual(gc.garbage, [])
 680                 rc = sys.gettotalrefcount()
 681                 writer = csv.writer(s)
 682                 for row in rows:
 683                     writer.writerow(row)
 684                 delta = rc-lastrc
 685                 lastrc = rc
 686             # if writer leaks during write, last delta should be 5 or more
 687             self.assertEqual(delta < 5, True)
 688
 689 # commented out for now - csv module doesn't yet support Unicode
 690 if 0:
 691     from StringIO import StringIO
 692     import csv
 693
 694     class TestUnicode(unittest.TestCase):
 695         def test_unicode_read(self):
 696             import codecs
 697             f = codecs.EncodedFile(StringIO("Martin von Löwis,"
 698                                             "Marc André Lemburg,"
 699                                             "Guido van Rossum,"
 700                                             "François Pinard\r\n"),
 701                                    data_encoding='iso-8859-1')
 702             reader = csv.reader(f)
 703             self.assertEqual(list(reader), [[u"Martin von Löwis",
 704                                              u"Marc André Lemburg",
 705                                              u"Guido van Rossum",
 706                                              u"François Pinardn"]])
 707
 708 def test_main():
 709     mod = sys.modules[__name__]
 710     test_support.run_unittest(
 711         *[getattr(mod, name) for name in dir(mod) if name.startswith('Test')]
 712     )
 713
 714 if __name__ == '__main__':
 715     test_main()