1 | """Mozilla / Netscape cookie loading / saving."""
|
---|
2 |
|
---|
3 | import re, time
|
---|
4 |
|
---|
5 | from cookielib import (_warn_unhandled_exception, FileCookieJar, LoadError,
|
---|
6 | Cookie, MISSING_FILENAME_TEXT)
|
---|
7 |
|
---|
8 | class MozillaCookieJar(FileCookieJar):
|
---|
9 | """
|
---|
10 |
|
---|
11 | WARNING: you may want to backup your browser's cookies file if you use
|
---|
12 | this class to save cookies. I *think* it works, but there have been
|
---|
13 | bugs in the past!
|
---|
14 |
|
---|
15 | This class differs from CookieJar only in the format it uses to save and
|
---|
16 | load cookies to and from a file. This class uses the Mozilla/Netscape
|
---|
17 | `cookies.txt' format. lynx uses this file format, too.
|
---|
18 |
|
---|
19 | Don't expect cookies saved while the browser is running to be noticed by
|
---|
20 | the browser (in fact, Mozilla on unix will overwrite your saved cookies if
|
---|
21 | you change them on disk while it's running; on Windows, you probably can't
|
---|
22 | save at all while the browser is running).
|
---|
23 |
|
---|
24 | Note that the Mozilla/Netscape format will downgrade RFC2965 cookies to
|
---|
25 | Netscape cookies on saving.
|
---|
26 |
|
---|
27 | In particular, the cookie version and port number information is lost,
|
---|
28 | together with information about whether or not Path, Port and Discard were
|
---|
29 | specified by the Set-Cookie2 (or Set-Cookie) header, and whether or not the
|
---|
30 | domain as set in the HTTP header started with a dot (yes, I'm aware some
|
---|
31 | domains in Netscape files start with a dot and some don't -- trust me, you
|
---|
32 | really don't want to know any more about this).
|
---|
33 |
|
---|
34 | Note that though Mozilla and Netscape use the same format, they use
|
---|
35 | slightly different headers. The class saves cookies using the Netscape
|
---|
36 | header by default (Mozilla can cope with that).
|
---|
37 |
|
---|
38 | """
|
---|
39 | magic_re = "#( Netscape)? HTTP Cookie File"
|
---|
40 | header = """\
|
---|
41 | # Netscape HTTP Cookie File
|
---|
42 | # http://www.netscape.com/newsref/std/cookie_spec.html
|
---|
43 | # This is a generated file! Do not edit.
|
---|
44 |
|
---|
45 | """
|
---|
46 |
|
---|
47 | def _really_load(self, f, filename, ignore_discard, ignore_expires):
|
---|
48 | now = time.time()
|
---|
49 |
|
---|
50 | magic = f.readline()
|
---|
51 | if not re.search(self.magic_re, magic):
|
---|
52 | f.close()
|
---|
53 | raise LoadError(
|
---|
54 | "%r does not look like a Netscape format cookies file" %
|
---|
55 | filename)
|
---|
56 |
|
---|
57 | try:
|
---|
58 | while 1:
|
---|
59 | line = f.readline()
|
---|
60 | if line == "": break
|
---|
61 |
|
---|
62 | # last field may be absent, so keep any trailing tab
|
---|
63 | if line.endswith("\n"): line = line[:-1]
|
---|
64 |
|
---|
65 | # skip comments and blank lines XXX what is $ for?
|
---|
66 | if (line.strip().startswith(("#", "$")) or
|
---|
67 | line.strip() == ""):
|
---|
68 | continue
|
---|
69 |
|
---|
70 | domain, domain_specified, path, secure, expires, name, value = \
|
---|
71 | line.split("\t")
|
---|
72 | secure = (secure == "TRUE")
|
---|
73 | domain_specified = (domain_specified == "TRUE")
|
---|
74 | if name == "":
|
---|
75 | # cookies.txt regards 'Set-Cookie: foo' as a cookie
|
---|
76 | # with no name, whereas cookielib regards it as a
|
---|
77 | # cookie with no value.
|
---|
78 | name = value
|
---|
79 | value = None
|
---|
80 |
|
---|
81 | initial_dot = domain.startswith(".")
|
---|
82 | assert domain_specified == initial_dot
|
---|
83 |
|
---|
84 | discard = False
|
---|
85 | if expires == "":
|
---|
86 | expires = None
|
---|
87 | discard = True
|
---|
88 |
|
---|
89 | # assume path_specified is false
|
---|
90 | c = Cookie(0, name, value,
|
---|
91 | None, False,
|
---|
92 | domain, domain_specified, initial_dot,
|
---|
93 | path, False,
|
---|
94 | secure,
|
---|
95 | expires,
|
---|
96 | discard,
|
---|
97 | None,
|
---|
98 | None,
|
---|
99 | {})
|
---|
100 | if not ignore_discard and c.discard:
|
---|
101 | continue
|
---|
102 | if not ignore_expires and c.is_expired(now):
|
---|
103 | continue
|
---|
104 | self.set_cookie(c)
|
---|
105 |
|
---|
106 | except IOError:
|
---|
107 | raise
|
---|
108 | except Exception:
|
---|
109 | _warn_unhandled_exception()
|
---|
110 | raise LoadError("invalid Netscape format cookies file %r: %r" %
|
---|
111 | (filename, line))
|
---|
112 |
|
---|
113 | def save(self, filename=None, ignore_discard=False, ignore_expires=False):
|
---|
114 | if filename is None:
|
---|
115 | if self.filename is not None: filename = self.filename
|
---|
116 | else: raise ValueError(MISSING_FILENAME_TEXT)
|
---|
117 |
|
---|
118 | f = open(filename, "w")
|
---|
119 | try:
|
---|
120 | f.write(self.header)
|
---|
121 | now = time.time()
|
---|
122 | for cookie in self:
|
---|
123 | if not ignore_discard and cookie.discard:
|
---|
124 | continue
|
---|
125 | if not ignore_expires and cookie.is_expired(now):
|
---|
126 | continue
|
---|
127 | if cookie.secure: secure = "TRUE"
|
---|
128 | else: secure = "FALSE"
|
---|
129 | if cookie.domain.startswith("."): initial_dot = "TRUE"
|
---|
130 | else: initial_dot = "FALSE"
|
---|
131 | if cookie.expires is not None:
|
---|
132 | expires = str(cookie.expires)
|
---|
133 | else:
|
---|
134 | expires = ""
|
---|
135 | if cookie.value is None:
|
---|
136 | # cookies.txt regards 'Set-Cookie: foo' as a cookie
|
---|
137 | # with no name, whereas cookielib regards it as a
|
---|
138 | # cookie with no value.
|
---|
139 | name = ""
|
---|
140 | value = cookie.name
|
---|
141 | else:
|
---|
142 | name = cookie.name
|
---|
143 | value = cookie.value
|
---|
144 | f.write(
|
---|
145 | "\t".join([cookie.domain, initial_dot, cookie.path,
|
---|
146 | secure, expires, name, value])+
|
---|
147 | "\n")
|
---|
148 | finally:
|
---|
149 | f.close()
|
---|