Test Suite Housekeeping (Part 3) (#1167)

* Add test fixtures exclusive for tests.providers

* Pytestify tests.sphinx

* Pytestify and improve user agent provider tests

* Pytestify barcode provider tests

* Fix flake8 bare except issue

* Pytestify, reorganize, and improve Generator class tests

* Pytestify, reorganize, and improve Faker class tests

* Fix test failure

* Pytestify currency provider tests

* Pytestify isbn provider tests

* Pytestify, reorganize, and improve internet provider tests

* Pytestify, reorganize, and improve base provider tests

* Fix CI errors

Author: malefice

tests/providers/__init__.py

@@ -1,53 +1,229 @@
-import unittest
+import re
+import string
 
-from faker import Faker
+from collections import OrderedDict
 
+import pytest
 
-class BaseProviderTestCase(unittest.TestCase):
-    def setUp(self):
-        self.fake = Faker('en_US')
-        Faker.seed(0)
+from faker.providers import BaseProvider
 
-    def test_random_digit_or_empty(self):
-        ret = self.fake.random_digit_or_empty()
-        assert isinstance(ret, int)
-        assert 0 <= ret <= 9
 
-        Faker.seed(1)
-        assert self.fake.random_digit_or_empty() == ''
+class TestBaseProvider:
+    """Test base provider methods"""
 
-    def test_random_digit_not_null_or_empty(self):
-        ret = self.fake.random_digit_not_null_or_empty()
-        assert isinstance(ret, int)
-        assert 0 <= ret <= 9
+    def test_locale(self, faker, num_samples):
+        locales = [
+            '{}_{}'.format(language, region)
+            for language, regions in BaseProvider.language_locale_codes.items()
+            for region in regions
+        ]
+        for _ in range(num_samples):
+            locale = faker.locale()
+            assert locale in locales
 
-        Faker.seed(1)
-        assert self.fake.random_digit_not_null_or_empty() == ''
+    def test_language_code(self, faker, num_samples):
+        language_codes = list(BaseProvider.language_locale_codes)
+        for _ in range(num_samples):
+            language_code = faker.language_code()
+            assert language_code in language_codes
 
-    def test_randomize_nb_elements(self):
-        assert self.fake.randomize_nb_elements(number=1, le=True, ge=True) == 1
+    def test_random_digit(self, faker, num_samples):
+        samples = [faker.random_digit() for _ in range(num_samples * 10)]
+        assert set(samples) == set(range(10))
 
-        assert self.fake.randomize_nb_elements(le=True, ge=True) == 10
+    def test_random_digit_not_null(self, faker, num_samples):
+        samples = [faker.random_digit_not_null() for _ in range(num_samples * 10)]
+        assert set(samples) == set(range(1, 10))
 
-        assert self.fake.randomize_nb_elements(min=42) == 42
-        assert self.fake.randomize_nb_elements(max=1) == 1
+    def test_random_digit_or_empty(self, faker, num_samples):
+        expected = set(range(10))
+        expected.add('')
+        samples = [faker.random_digit_or_empty() for _ in range(num_samples * 10)]
+        assert set(samples) == expected
+
+    def test_random_digit_not_null_or_empty(self, faker, num_samples):
+        expected = set(range(1, 10))
+        expected.add('')
+        samples = [faker.random_digit_not_null_or_empty() for _ in range(num_samples * 10)]
+        assert set(samples) == expected
+
+    def test_random_number(self, faker):
+        number = faker.random_number(10, True)
+        assert len(str(number)) == 10
+
+        # Digits parameter < 0
+        with pytest.raises(ValueError):
+            number = faker.random_number(-1, True)
+
+        # Digits parameter < 1 with fix_len=True
+        with pytest.raises(ValueError):
+            number = faker.random_number(0, True)
+
+    @pytest.mark.parametrize('text,pattern', [
+        ('', r''),
+        ('abcd', r'abcd'),
+        ('#' * 100, r'[0-9]{100}'),
+        ('%' * 100, r'[1-9]{100}'),
+        ('!' * 100, r'[0-9]{,100}'),
+        ('@' * 100, r'[0-9]{,100}'),
+        ('##!abc %%@def##!' * 100, r'(?:[0-9]{2,3}abc [1-9]{2,3}def[0-9]{2,3}){100}'),
+        ('#@@#^?あ5漢!!%%@' * 100, r'(?:\d[1-9]{,2}\d\^\?あ5漢\d{,2}[1-9]{2}[1-9]*){100}'),
+    ], ids=[
+        'empty_string',
+        'no_valid_placeholders',
+        'only_number_signs',
+        'only_percent_signs',
+        'only_exclamation_marks',
+        'only_at_symbols',
+        'with_ascii_characters',
+        'with_other_symbols_and_non_ascii',
+    ])
+    def test_numerify(self, faker, num_samples, text, pattern):
+        for _ in range(num_samples):
+            numerified = faker.numerify(text)
+            assert re.fullmatch(pattern, numerified)
+
+    @pytest.mark.parametrize('text,letters,pattern', [
+        ('', string.ascii_letters, r''),
+        ('abcd', string.ascii_letters, r'abcd'),
+        ('???', string.ascii_letters, r'[0-9a-zA-Z]{3}'),
+        ('???', 'aBcDeFgHiJ12345', r'[1-5aBcDeFgHiJ]{3}'),
+        ('??Xr^#7p??', 'AbCdخあ5漢7Я', r'[AbCdخあ5漢7Я]{2}Xr\^#7p[AbCdخあ5漢7Я]{2}'),
+    ], ids=[
+        'empty_string',
+        'no_valid_placeholders',
+        'letters_using_whole_ascii',
+        'letters_using_ascii_subset',
+        'pattern_with_other_symbols_and_letters_using_non_ascii',
+    ])
+    def test_lexify(self, faker, num_samples, text, letters, pattern):
+        for _ in range(num_samples):
+            lexified = faker.lexify(text, letters=letters)
+            assert re.fullmatch(pattern, lexified)
+
+    @pytest.mark.parametrize('text,letters,pattern', [
+        ('', string.ascii_letters, r''),
+        ('abcd', string.ascii_letters, r'abcd'),
+        ('???', string.ascii_letters, r'[0-9a-zA-Z]{3}'),
+        ('???', 'aBcDeFgHiJ12345', r'[1-5aBcDeFgHiJ]{3}'),
+        ('#%!@???', string.ascii_letters, r'\d[1-9]\d*[1-9]*[0-9a-zA-Z]{3}'),
+        ('#%!@???', 'aBcDeFgHiJ12345', r'\d[1-9]\d*[1-9]*[1-5aBcDeFgHiJ]{3}'),
+        ('#%!@??Xr7p??', 'AbCdخあ5漢7Я', r'\d[1-9]\d*[1-9]*[AbCdخあ5漢7Я]{2}Xr7p[AbCdخあ5漢7Я]{2}'),
+    ], ids=[
+        'empty_string',
+        'no_valid_placeholders',
+        'simple_pattern_and_letters_using_whole_ascii',
+        'simple_pattern_and_letters_using_ascii_subset',
+        'more_complex_pattern_and_letters_using_whole_ascii',
+        'more_complex_pattern_and_letters_using_ascii_subset',
+        'more_complex_pattern_with_other_symbols_and_letters_using_non_ascii',
+    ])
+    def test_bothify(self, faker, num_samples, text, letters, pattern):
+        for _ in range(num_samples):
+            bothified = faker.bothify(text, letters=letters)
+            assert re.fullmatch(pattern, bothified)
+
+    @pytest.mark.parametrize('text,upper,pattern', [
+        ('', False, r''),
+        ('', True, r''),
+        ('abcd', False, r'abcd'),
+        ('abcd', True, r'abcd'),
+        ('^^^^', False, r'[0-9a-f]{4}'),
+        ('^^^^', True, r'[0-9A-F]{4}'),
+        ('Abc ^^^ %^^^?あ5漢!#^^', False, r'Abc [0-9a-f]{3} %[0-9a-f]{3}\?あ5漢!#[0-9a-f]{2}'),
+        ('Abc ^^^ %^^^?あ5漢!#^^', True, r'Abc [0-9A-F]{3} %[0-9A-F]{3}\?あ5漢!#[0-9A-F]{2}'),
+    ], ids=[
+        'empty_string_lowercase',
+        'empty_string_uppercase',
+        'no_circumflex_lowercase',
+        'no_circumflex_uppercase',
+        'simple_pattern_lowercase',
+        'simple_pattern_uppercase',
+        'complex_pattern_lowercase',
+        'complex_pattern_uppercase',
+    ])
+    def test_hexify(self, faker, num_samples, text, upper, pattern):
+        for _ in range(num_samples):
+            hexified = faker.hexify(text, upper=upper)
+            assert re.fullmatch(pattern, hexified)
+
+    def test_random_letter(self, faker, num_samples):
+        for _ in range(num_samples):
+            letter = faker.random_letter()
+            assert letter.isalpha()
+
+    def test_random_lowercase_letter(self, faker, num_samples):
+        for _ in range(num_samples):
+            letter = faker.random_lowercase_letter()
+            assert letter.isalpha() and letter.lower() == letter
+
+    def test_random_uppercase_letter(self, faker, num_samples):
+        for _ in range(num_samples):
+            letter = faker.random_uppercase_letter()
+            assert letter.isalpha() and letter.upper() == letter
+
+    def test_random_element(self, faker, num_samples):
+        # dicts not allowed because they introduce dependency on PYTHONHASHSEED
+        with pytest.raises(ValueError):
+            faker.random_element({})
+
+        choices = ('a', 'b', 'c', 'd')
+        for _ in range(num_samples):
+            assert faker.random_element(choices) in choices
+
+        choices = OrderedDict([('a', 5), ('b', 2), ('c', 2), ('d', 1)])
+        for _ in range(num_samples):
+            assert faker.random_element(choices) in choices
+
+        choices = OrderedDict([('a', 0.5), ('b', 0.2), ('c', 0.2), ('d', 0.1)])
+        for _ in range(num_samples):
+            assert faker.random_element(choices) in choices
+
+    def test_random_sample(self, faker):
+        # Too many items requested
+        with pytest.raises(ValueError):
+            faker.random_sample('abcde', 6)
+
+        # Same length
+        sample = faker.random_sample('abcd', 4)
+        assert sorted(sample) == list('abcd')
+
+        sample = faker.random_sample('abcde', 5)
+        assert sorted(sample) == list('abcde')
+
+        # Length = 3
+        sample = faker.random_sample('abcde', 3)
+        assert len(sample) == 3
+        assert set(sample).issubset(set('abcde'))
+
+        # Length = 1
+        sample = faker.random_sample('abcde', 1)
+        assert len(sample) == 1
+        assert set(sample).issubset(set('abcde'))
+
+        # Length = 0
+        sample = faker.random_sample('abcde', 0)
+        assert sample == []
+
+    def test_randomize_nb_elements(self, faker, num_samples):
+        assert faker.randomize_nb_elements(number=1, le=True, ge=True) == 1
+        assert faker.randomize_nb_elements(le=True, ge=True) == 10
+        assert faker.randomize_nb_elements(min=42) == 42
+        assert faker.randomize_nb_elements(max=1) == 1
 
         number = 9999
-        random_times = 100
         lower_bound = int(number * 0.6)
         upper_bound = int(number * 1.4)
 
-        for _ in range(random_times):
-            res = self.fake.randomize_nb_elements(number=number, le=True)
+        for _ in range(num_samples):
+            res = faker.randomize_nb_elements(number=number, le=True)
             assert res >= lower_bound
             assert res <= number, "'{}' is not <= than '{}'".format(res, number)
 
-        for _ in range(random_times):
-            res = self.fake.randomize_nb_elements(number=number, ge=True)
-            assert res >= number
-            assert res <= upper_bound
+        for _ in range(num_samples):
+            res = faker.randomize_nb_elements(number=number, ge=True)
+            assert number <= res <= upper_bound
 
-        for _ in range(random_times):
-            res = self.fake.randomize_nb_elements(number=number)
-            assert res >= lower_bound
-            assert res <= upper_bound
+        for _ in range(num_samples):
+            res = faker.randomize_nb_elements(number=number)
+            assert lower_bound <= res <= upper_bound

tests/providers/conftest.py

@@ -0,0 +1,41 @@
+import locale as pylocale
+import re
+
+import pytest
+
+from faker import Faker
+from faker.contrib.pytest.plugin import DEFAULT_SEED
+
+LOCALE_TEST_CLASS_NAME_REGEX = re.compile(
+    r'^Test(?P<language>[A-Z][a-z]{1,2})(?P<region>[A-Z][a-z])$',
+)
+
+
+@pytest.fixture(scope='class', autouse=True)
+def _class_locale_faker(request):
+    if not request.cls:
+        return None
+    class_name = request.cls.__name__
+    match = LOCALE_TEST_CLASS_NAME_REGEX.fullmatch(class_name)
+    if not match:
+        return None
+    locale = '{language}_{region}'.format(**match.groupdict())
+    locale = pylocale.normalize(locale).split('.')[0]
+    return Faker(locale=locale)
+
+
+@pytest.fixture(autouse=True)
+def faker(_class_locale_faker, faker):
+    if not _class_locale_faker:
+        return faker
+    _class_locale_faker.seed_instance(DEFAULT_SEED)
+    return _class_locale_faker
+
+
+@pytest.fixture(scope='class', autouse=True)
+def num_samples(request):
+    try:
+        num = int(request.cls.num_samples)
+    except AttributeError:
+        num = 100
+    return num