Example #1
0
    def test_get_transliteration_rules(self):
        self.config_rules()
        loader = ICURuleLoader(self.project_env)
        rules = loader.get_transliteration_rules()
        trans = Transliterator.createFromRules("test", rules)

        assert trans.transliterate(" проспект-Prospekt ") == " prospekt Prospekt "
Example #2
0
    def test_empty_rule_set(self):
        self.write_config("""\
            normalization:
            transliteration:
            token-analysis:
              - analyzer: generic
                variants:
            """)

        rules = ICURuleLoader(self.project_env)
        assert rules.get_search_rules() == ''
        assert rules.get_normalization_rules() == ''
        assert rules.get_transliteration_rules() == ''
Example #3
0
def test_empty_rule_set(test_config):
    (test_config.project_dir / 'icu_tokenizer.yaml').write_text(
        dedent("""\
        normalization:
        transliteration:
        token-analysis:
          - analyzer: generic
            variants:
        """))

    rules = ICURuleLoader(test_config)
    assert rules.get_search_rules() == ''
    assert rules.get_normalization_rules() == ''
    assert rules.get_transliteration_rules() == ''
Example #4
0
    def test_transliteration_rules_from_file(self):
        self.write_config("""\
            normalization:
            transliteration:
                - "'ax' > 'b'"
                - !include transliteration.yaml
            token-analysis:
                - analyzer: generic
                  variants:
            """)
        transpath = self.project_env.project_dir / ('transliteration.yaml')
        transpath.write_text('- "x > y"')

        loader = ICURuleLoader(self.project_env)
        rules = loader.get_transliteration_rules()
        trans = Transliterator.createFromRules("test", rules)

        assert trans.transliterate(" axxt ") == " byt "
Example #5
0
def test_transliteration_rules_from_file(test_config):
    cfgpath = test_config.project_dir / ('icu_tokenizer.yaml')
    cfgpath.write_text(
        dedent("""\
        normalization:
        transliteration:
            - "'ax' > 'b'"
            - !include transliteration.yaml
        token-analysis:
            - analyzer: generic
              variants:
        """))
    transpath = test_config.project_dir / ('transliteration.yaml')
    transpath.write_text('- "x > y"')

    loader = ICURuleLoader(test_config)
    rules = loader.get_transliteration_rules()
    trans = Transliterator.createFromRules("test", rules)

    assert trans.transliterate(" axxt ") == " byt "
Example #6
0
def test_get_transliteration_rules(cfgrules):
    loader = ICURuleLoader(cfgrules())
    rules = loader.get_transliteration_rules()
    trans = Transliterator.createFromRules("test", rules)

    assert trans.transliterate(" проспект-Prospekt ") == " prospekt Prospekt "