import pytest

from protego import Protego


@pytest.mark.parametrize(
    'path,user_agent',
    [
        ('/group1', 'hedwig-news'),
        ('/group1', 'hedwig-news2'),
        ('/group2', 'hedwi'),
        ('/group2', 'a'),
        ('/group3', 'hedwig-new'),
        ('/group3', 'hedwig-images'),
    ]
)
def test_user_agent_precedence(path, user_agent):
    robotstxt_content = u"""
    user-agent: hedwig-news
    disallow: /
    allow: /group1
    user-agent: *
    disallow: /
    allow: /group2
    user-agent: hedwig
    disallow: /
    allow: /group3
    """
    rp = Protego.parse(content=robotstxt_content)
    for allowed_path in ('/group1', '/group2', '/group3'):
        if rp.can_fetch(allowed_path, user_agent):
            break
    else:
        allowed_path = None
    assert allowed_path == path


@pytest.mark.parametrize(
    'pattern,path,match',
    [
        ('/', '/harry', True),
        ('/', '/device/time-turner', True),
        ('/', '/hogwards.html', True),
        ('/*', '/harry', True),
        ('/*', '/device/time-turner', True),
        ('/*', '/hogwards.html', True),

        ('/phoenix', '/phoenix', True),
        ('/phoenix', '/phoenix.html', True),
        ('/phoenix', '/phoenix/sparky.html', True),
        ('/phoenix', '/phoenixheads', True),
        ('/phoenix', '/phoenixheads/yummy.html', True),
        ('/phoenix', '/phoenix.php?id=anything', True),
        ('/phoenix', '/Phoenix.asp', False),
        ('/phoenix', '/redphoenix', False),
        ('/phoenix', '/?id=phoenix', False),

        ('/phoenix*', '/phoenix', True),
        ('/phoenix*', '/phoenix.html', True),
        ('/phoenix*', '/phoenix/sparky.html', True),
        ('/phoenix*', '/phoenixheads', True),
        ('/phoenix*', '/phoenixheads/yummy.html', True),
        ('/phoenix*', '/phoenix.php?id=anything', True),
        ('/phoenix*', '/Phoenix.asp', False),
        ('/phoenix*', '/redphoenix', False),
        ('/phoenix*', '/?id=phoenix', False),

        ('/phoenix/', '/phoenix/', True),
        ('/phoenix/', '/phoenix/?id=anything', True),
        ('/phoenix/', '/phoenix/sparky.htm', True),
        ('/phoenix/', '/phoenix', False),
        ('/phoenix/', '/phoenix.html', False),
        ('/phoenix/', '/Phoenix/Sparky.asp', False),

        ('/*.php', '/filename.php', True),
        ('/*.php', '/folder/filename.php', True),
        ('/*.php', '/folder/filename.php?parameters', True),
        ('/*.php', '/folder/any.php.file.html', True),
        ('/*.php', '/filename.php/', True),
        ('/*.php', '/windows.PHP', False),
        ('/*.php', '/', False),
        ('/*.php', '/index?f=filename.php/', True),
        ('/*.php', '/index?php', False),

        ('/*.php$', '/filename.php', True),
        ('/*.php$', '/folder/filename.php', True),
        ('/*.php$', '/filename.php?parameters', False),
        ('/*.php$', '/filename.php/', False),
        ('/*.php$', '/filename.php5', False),
        ('/*.php$', '/windows.PHP', False),
        ('/*.php$', '/filename?php', False),

        ('/fish*.php', '/fish.php', True),
        ('/fish*.php', '/fishheads/catfish.php?parameters', True),
        ('/fish*.php', '/Fish.PHP', False),

    ]
)
def test_path_matching(pattern, path, match):
    content = """
    User-Agent: *
    disallow: {}
    """.format(pattern)
    rp = Protego.parse(content)
    assert (not rp.can_fetch(path, '*')) == match


@pytest.mark.parametrize(
    'rules,url,allowed',
    [
        ("allow: /p \n disallow: /", "http://example.com/page", True),
        ("allow: /folder \n disallow: /folder",
         "http://example.com/folder/page", True),
        ("allow: /$ \n disallow: /", "http://example.com/", True),
        ("allow: /$ \n disallow: /", "http://example.com/page.htm", False),
    ]
)
def test_record_precedence(rules, url, allowed):
    content = """
    User-Agent: *
    {}
    """.format(rules)
    rp = Protego.parse(content)
    assert rp.can_fetch(url, '*') == allowed