"""Tests for http/cookiejar.py."""

import os
import re
import test.support
import time
import unittest
import urllib.request
import pathlib

from http.cookiejar import (time2isoz, http2time, iso2time, time2netscape,
     parse_ns_headers, join_header_words, split_header_words, Cookie,
     CookieJar, DefaultCookiePolicy, LWPCookieJar, MozillaCookieJar,
     LoadError, lwp_cookie_str, DEFAULT_HTTP_PORT, escape_path,
     reach, is_HDN, domain_match, user_domain_match, request_path,
     request_port, request_host)


class DateTimeTests(unittest.TestCase):

    def test_time2isoz(self):
        base = 1019227000
        day = 24*3600
        self.assertEqual(time2isoz(base), "2002-04-19 14:36:40Z")
        self.assertEqual(time2isoz(base+day), "2002-04-20 14:36:40Z")
        self.assertEqual(time2isoz(base+2*day), "2002-04-21 14:36:40Z")
        self.assertEqual(time2isoz(base+3*day), "2002-04-22 14:36:40Z")

        az = time2isoz()
        bz = time2isoz(500000)
        for text in (az, bz):
            self.assertRegex(text, r"^\d{4}-\d\d-\d\d \d\d:\d\d:\d\dZ$",
                             "bad time2isoz format: %s %s" % (az, bz))

    def test_time2netscape(self):
        base = 1019227000
        day = 24*3600
        self.assertEqual(time2netscape(base), "Fri, 19-Apr-2002 14:36:40 GMT")
        self.assertEqual(time2netscape(base+day),
                         "Sat, 20-Apr-2002 14:36:40 GMT")

        self.assertEqual(time2netscape(base+2*day),
                         "Sun, 21-Apr-2002 14:36:40 GMT")

        self.assertEqual(time2netscape(base+3*day),
                         "Mon, 22-Apr-2002 14:36:40 GMT")

        az = time2netscape()
        bz = time2netscape(500000)
        for text in (az, bz):
            # Format "%s, %02d-%s-%04d %02d:%02d:%02d GMT"
            self.assertRegex(
                text,
                r"[a-zA-Z]{3}, \d{2}-[a-zA-Z]{3}-\d{4} \d{2}:\d{2}:\d{2} GMT$",
                "bad time2netscape format: %s %s" % (az, bz))

    def test_http2time(self):
        def parse_date(text):
            return time.gmtime(http2time(text))[:6]

        self.assertEqual(parse_date("01 Jan 2001"), (2001, 1, 1, 0, 0, 0.0))

        # this test will break around year 2070
        self.assertEqual(parse_date("03-Feb-20"), (2020, 2, 3, 0, 0, 0.0))

        # this test will break around year 2048
        self.assertEqual(parse_date("03-Feb-98"), (1998, 2, 3, 0, 0, 0.0))

    def test_http2time_formats(self):
        # test http2time for supported dates.  Test cases with 2 digit year
        # will probably break in year 2044.
        tests = [
         'Thu, 03 Feb 1994 00:00:00 GMT',  # proposed new HTTP format
         'Thursday, 03-Feb-94 00:00:00 GMT',  # old rfc850 HTTP format
         'Thursday, 03-Feb-1994 00:00:00 GMT',  # broken rfc850 HTTP format

         '03 Feb 1994 00:00:00 GMT',  # HTTP format (no weekday)
         '03-Feb-94 00:00:00 GMT',  # old rfc850 (no weekday)
         '03-Feb-1994 00:00:00 GMT',  # broken rfc850 (no weekday)
         '03-Feb-1994 00:00 GMT',  # broken rfc850 (no weekday, no seconds)
         '03-Feb-1994 00:00',  # broken rfc850 (no weekday, no seconds, no tz)
         '02-Feb-1994 24:00',  # broken rfc850 (no weekday, no seconds,
                               # no tz) using hour 24 with yesterday date

         '03-Feb-94',  # old rfc850 HTTP format (no weekday, no time)
         '03-Feb-1994',  # broken rfc850 HTTP format (no weekday, no time)
         '03 Feb 1994',  # proposed new HTTP format (no weekday, no time)

         # A few tests with extra space at various places
         '  03   Feb   1994  0:00  ',
         '  03-Feb-1994  ',
        ]

        test_t = 760233600  # assume broken POSIX counting of seconds
        result = time2isoz(test_t)
        expected = "1994-02-03 00:00:00Z"
        self.assertEqual(result, expected,
                         "%s  =>  '%s' (%s)" % (test_t, result, expected))

        for s in tests:
            self.assertEqual(http2time(s), test_t, s)
            self.assertEqual(http2time(s.lower()), test_t, s.lower())
            self.assertEqual(http2time(s.upper()), test_t, s.upper())

    def test_http2time_garbage(self):
        for test in [
            '',
            'Garbage',
            'Mandag 16. September 1996',
            '01-00-1980',
            '01-13-1980',
            '00-01-1980',
            '32-01-1980',
            '01-01-1980 25:00:00',
            '01-01-1980 00:61:00',
            '01-01-1980 00:00:62',
            '08-Oct-3697739',
            '08-01-3697739',
            '09 Feb 19942632 22:23:32 GMT',
            'Wed, 09 Feb 1994834 22:23:32 GMT',
            ]:
            self.assertIsNone(http2time(test),
                              "http2time(%s) is not None\n"
                              "http2time(test) %s" % (test, http2time(test)))

    def test_http2time_redos_regression_actually_completes(self):
        # LOOSE_HTTP_DATE_RE was vulnerable to malicious input which caused catastrophic backtracking (REDoS).
        # If we regress to cubic complexity, this test will take a very long time to succeed.
        # If fixed, it should complete within a fraction of a second.
        http2time("01 Jan 1970{}00:00:00 GMT!".format(" " * 10 ** 5))
        http2time("01 Jan 1970 00:00:00{}GMT!".format(" " * 10 ** 5))

    def test_iso2time(self):
        def parse_date(text):
            return time.gmtime(iso2time(text))[:6]

        # ISO 8601 compact format
        self.assertEqual(parse_date("19940203T141529Z"),
                         (1994, 2, 3, 14, 15, 29))

        # ISO 8601 with time behind UTC
        self.assertEqual(parse_date("1994-02-03 07:15:29 -0700"),
                         (1994, 2, 3, 14, 15, 29))

        # ISO 8601 with time ahead of UTC
        self.assertEqual(parse_date("1994-02-03 19:45:29 +0530"),
                         (1994, 2, 3, 14, 15, 29))

    def test_iso2time_formats(self):
        # test iso2time for supported dates.
        tests = [
            '1994-02-03 00:00:00 -0000', # ISO 8601 format
            '1994-02-03 00:00:00 +0000', # ISO 8601 format
            '1994-02-03 00:00:00',       # zone is optional
            '1994-02-03',                # only date
            '1994-02-03T00:00:00',       # Use T as separator
            '19940203',                  # only date
            '1994-02-02 24:00:00',       # using hour-24 yesterday date
            '19940203T000000Z',          # ISO 8601 compact format

            # A few tests with extra space at various places
            '  1994-02-03 ',
            '  1994-02-03T00:00:00  ',
        ]

        test_t = 760233600  # assume broken POSIX counting of seconds
        for s in tests:
            self.assertEqual(iso2time(s), test_t, s)
            self.assertEqual(iso2time(s.lower()), test_t, s.lower())
            self.assertEqual(iso2time(s.upper()), test_t, s.upper())

    def test_iso2time_garbage(self):
        for test in [
            '',
            'Garbage',
            'Thursday, 03-Feb-94 00:00:00 GMT',
            '1980-00-01',
            '1980-13-01',
            '1980-01-00',
            '1980-01-32',
            '1980-01-01 25:00:00',
            '1980-01-01 00:61:00',
            '01-01-1980 00:00:62',
            '01-01-1980T00:00:62',
            '19800101T250000Z',
            ]:
            self.assertIsNone(iso2time(test),
                              "iso2time(%r)" % test)

    def test_iso2time_performance_regression(self):
        # If ISO_DATE_RE regresses to quadratic complexity, this test will take a very long time to succeed.
        # If fixed, it should complete within a fraction of a second.
        iso2time('1994-02-03{}14:15:29 -0100!'.format(' '*10**6))
        iso2time('1994-02-03 14:15:29{}-0100!'.format(' '*10**6))


class HeaderTests(unittest.TestCase):

    def test_parse_ns_headers(self):
        # quotes should be stripped
        expected = [[('foo', 'bar'), ('expires', 2209069412), ('version', '0')]]
        for hdr in [
            'foo=bar; expires=01 Jan 2040 22:23:32 GMT',
            'foo=bar; expires="01 Jan 2040 22:23:32 GMT"',
            ]:
            self.assertEqual(parse_ns_headers([hdr]), expected)

    def test_parse_ns_headers_version(self):

        # quotes should be stripped
        expected = [[('foo', 'bar'), ('version', '1')]]
        for hdr in [
            'foo=bar; version="1"',
            'foo=bar; Version="1"',
            ]:
            self.assertEqual(parse_ns_headers([hdr]), expected)

    def test_parse_ns_headers_special_names(self):
        # names such as 'expires' are not special in first name=value pair
        # of Set-Cookie: header
        # Cookie with name 'expires'
        hdr = 'expires=01 Jan 2040 22:23:32 GMT'
        expected = [[("expires", "01 Jan 2040 22:23:32 GMT"), ("version", "0")]]
        self.assertEqual(parse_ns_headers([hdr]), expected)

    def test_join_header_words(self):
        joined = join_header_words([[("foo", None), ("bar", "baz")]])
        self.assertEqual(joined, "foo; bar=baz")

        self.assertEqual(join_header_words([[]]), "")

    def test_split_header_words(self):
        tests = [
            ("foo", [[("foo", None)]]),
            ("foo=bar", [[("foo", "bar")]]),
            ("   foo   ", [[("foo", None)]]),
            ("   foo=   ", [[("foo", "")]]),
            ("   foo=", [[("foo", "")]]),
            ("   foo=   ; ", [[("foo", "")]]),
            ("   foo=   ; bar= baz ", [[("foo", ""), ("bar", "baz")]]),
            ("foo=bar bar=baz", [[("foo", "bar"), ("bar", "baz")]]),
            # doesn't really matter if this next fails, but it works ATM
            ("foo= bar=baz", [[("foo", "bar=baz")]]),
            ("foo=bar;bar=baz", [[("foo", "bar"), ("bar", "baz")]]),
            ('foo bar baz', [[("foo", None), ("bar", None), ("baz", None)]]),
            ("a, b, c", [[("a", None)], [("b", None)], [("c", None)]]),
            (r'foo; bar=baz, spam=, foo="\,\;\"", bar= ',
             [[("foo", None), ("bar", "baz")],
              [("spam", "")], [("foo", ',;"')], [("bar", "")]]),
            ]

        for arg, expect in tests:
            try:
                result = split_header_words([arg])
            except:
                import traceback, io
                f = io.StringIO()
                traceback.print_exc(None, f)
                result = "(error -- traceback follows)\n\n%s" % f.getvalue()
            self.assertEqual(result,  expect, """
When parsing: '%s'
Expected:     '%s'
Got:          '%s'
""" % (arg, expect, result))

    def test_roundtrip(self):
        tests = [
            ("foo", "foo"),
            ("foo=bar", "foo=bar"),
            ("   foo   ", "foo"),
            ("foo=", 'foo=""'),
            ("foo=bar bar=baz", "foo=bar; bar=baz"),
            ("foo=bar;bar=baz", "foo=bar; bar=baz"),
            ('foo bar baz', "foo; bar; baz"),
            (r'foo="\"" bar="\\"', r'foo="\""; bar="\\"'),
            ('foo,,,bar', 'foo, bar'),
            ('foo=bar,bar=baz', 'foo=bar, bar=baz'),

            ('text/html; charset=iso-8859-1',
             'text/html; charset="iso-8859-1"'),

            ('foo="bar"; port="80,81"; discard, bar=baz',
             'foo=bar; port="80,81"; discard, bar=baz'),

            (r'Basic realm="\"foo\\\\bar\""',
             r'Basic; realm="\"foo\\\\bar\""')
            ]

        for arg, expect in tests:
            input = split_header_words([arg])
            res = join_header_words(input)
            self.assertEqual(res, expect, """
When parsing: '%s'
Expected:     '%s'
Got:          '%s'
Input was:    '%s'
""" % (arg, expect, res, input))


class FakeResponse:
    def __init__(self, headers=[], url=None):
        """
        headers: list of RFC822-style 'Key: value' strings
        """
        import email
        self._headers = email.message_from_string("\n".join(headers))
        self._url = url
    def info(self): return self._headers

def interact_2965(cookiejar, url, *set_cookie_hdrs):
    return _interact(cookiejar, url, set_cookie_hdrs, "Set-Cookie2")

def interact_netscape(cookiejar, url, *set_cookie_hdrs):
    return _interact(cookiejar, url, set_cookie_hdrs, "Set-Cookie")

def _interact(cookiejar, url, set_cookie_hdrs, hdr_name):
    """Perform a single request / response cycle, returning Cookie: header."""
    req = urllib.request.Request(url)
    cookiejar.add_cookie_header(req)
    cookie_hdr = req.get_header("Cookie", "")
    headers = []
    for hdr in set_cookie_hdrs:
        headers.append("%s: %s" % (hdr_name, hdr))
    res = FakeResponse(headers, url)
    cookiejar.extract_cookies(res, req)
    return cookie_hdr


class FileCookieJarTests(unittest.TestCase):
    def test_constructor_with_str(self):
        filename = test.support.TESTFN
        c = LWPCookieJar(filename)
        self.assertEqual(c.filename, filename)

    def test_constructor_with_path_like(self):
        filename = pathlib.Path(test.support.TESTFN)
        c = LWPCookieJar(filename)
        self.assertEqual(c.filename, os.fspath(filename))

    def test_constructor_with_none(self):
        c = LWPCookieJar(None)
        self.assertIsNone(c.filename)

    def test_constructor_with_other_types(self):
        class A: