Mirror
/
cpython
mirror of https://github.com/python/cpython
4
1
Fork 0
Code Issues Releases Wiki Activity

Patch #963318: Add support for client-side cookie management.

This commit is contained in:
Martin v. Löwis 2004-05-31 18:22:40 +00:00
parent 0a6d0ff8d9
commit 2a6ba9097e
12 changed files with 4667 additions and 69 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
Doc/lib/lib.tex
View File

@ -233,6 +233,7 @@ and how to embed it in other applications.
\input{libbasehttp}
\input{libsimplehttp}
\input{libcgihttp}
\input{libcookielib}
\input{libcookie}
\input{libxmlrpclib}
\input{libsimplexmlrpc}

4
Doc/lib/libcookie.tex
View File

@ -68,6 +68,10 @@ you should not use the \class{SerialCookie} class.
\begin{seealso}
\seemodule{cookielib}{HTTP cookie handling for for web
\emph{clients}. The \module{cookielib} and \module{Cookie}
modules do not depend on each other.}
\seerfc{2109}{HTTP State Management Mechanism}{This is the state
management specification implemented by this module.}
\end{seealso}

679
Doc/lib/libcookielib.tex Normal file
View File

@ -0,0 +1,679 @@
\section{\module{cookielib} ---
Cookie handling for HTTP clients}
\declaremodule{standard}{cookielib}
\moduleauthor{John J. Lee}{jjl@pobox.com}
\sectionauthor{John J. Lee}{jjl@pobox.com}
\modulesynopsis{Cookie handling for HTTP clients}
The \module{cookielib} module defines classes for automatic handling
of HTTP cookies. It is useful for accessing web sites that require
small pieces of data -- \dfn{cookies} -- to be set on the client
machine by an HTTP response from a web server, and then returned to
the server in later HTTP requests.
Both the regular Netscape cookie protocol and the protocol defined by
\rfc{2965} are handled. RFC 2965 handling is switched off by default.
\rfc{2109} cookies are parsed as Netscape cookies and subsequently
treated as RFC 2965 cookies. Note that the great majority of cookies
on the Internet are Netscape cookies. \module{cookielib} attempts to
follow the de-facto Netscape cookie protocol (which differs
substantially from that set out in the original Netscape
specification), including taking note of the \code{max-age} and
\code{port} cookie-attributes introduced with RFC 2109. \note{The
various named parameters found in \mailheader{Set-Cookie} and
\mailheader{Set-Cookie2} headers (eg. \code{domain} and
\code{expires}) are conventionally referred to as \dfn{attributes}.
To distinguish them from Python attributes, the documentation for this
module uses the term \dfn{cookie-attribute} instead}.
The module defines the following exception:
\begin{excdesc}{LoadError}
Instances of \class{FileCookieJar} raise this exception on failure to
load cookies from a file.
\end{excdesc}
The following classes are provided:
\begin{classdesc}{CookieJar}{policy=\constant{None}}
\var{policy} is an object implementing the \class{CookiePolicy}
interface.
The \class{CookieJar} class stores HTTP cookies. It extracts cookies
from HTTP requests, and returns them in HTTP responses.
\class{CookieJar} instances automatically expire contained cookies
when necessary. Subclasses are also responsible for storing and
retrieving cookies from a file or database.
\end{classdesc}
\begin{classdesc}{FileCookieJar}{filename, delayload=\constant{None},
policy=\constant{None}}
\var{policy} is an object implementing the \class{CookiePolicy}
interface. For the other arguments, see the documentation for the
corresponding attributes.
A \class{CookieJar} which can load cookies from, and perhaps save
cookies to, a file on disk. Cookies are \strong{NOT} loaded from the
named file until either the \method{load()} or \method{revert()}
method is called. Subclasses of this class are documented in section
\ref{file-cookie-jar-classes}.
\end{classdesc}
\begin{classdesc}{CookiePolicy}{}
This class is responsible for deciding whether each cookie should be
accepted from / returned to the server.
\end{classdesc}
\begin{classdesc}{DefaultCookiePolicy}{
blocked_domains=\constant{None},
allowed_domains=\constant{None},
netscape=\constant{True}, rfc2965=\constant{False},
hide_cookie2=\constant{False},
strict_domain=\constant{False},
strict_rfc2965_unverifiable=\constant{True},
strict_ns_unverifiable=\constant{False},
strict_ns_domain=\constant{DefaultCookiePolicy.DomainLiberal},
strict_ns_set_initial_dollar=\constant{False},
strict_ns_set_path=\constant{False}
}
Constructor arguments should be passed as keyword arguments only.
\var{blocked_domains} is a sequence of domain names that we never
accept cookies from, nor return cookies to. \var{allowed_domains} if
not \constant{None}, this is a sequence of the only domains for which
we accept and return cookies. For all other arguments, see the
documentation for \class{CookiePolicy} and \class{DefaultCookiePolicy}
objects.
\class{DefaultCookiePolicy} implements the standard accept / reject
rules for Netscape and RFC 2965 cookies. RFC 2109 cookies
(ie. cookies received in a \mailheader{Set-Cookie} header with a
version cookie-attribute of 1) are treated according to the RFC 2965
rules. \class{DefaultCookiePolicy} also provides some parameters to
allow some fine-tuning of policy.
\end{classdesc}
\begin{classdesc}{Cookie}{}
This class represents Netscape, RFC 2109 and RFC 2965 cookies. It is
not expected that users of \module{cookielib} construct their own
\class{Cookie} instances. Instead, if necessary, call
\method{make_cookies()} on a \class{CookieJar} instance.
\end{classdesc}
\begin{seealso}
\seemodule{urllib2}{URL opening with automatic cookie handling.}
\seemodule{Cookie}{HTTP cookie classes, principally useful for
server-side code. The \module{cookielib} and \module{Cookie} modules
do not depend on each other.}
\seeurl{http://wwwsearch.sf.net/ClientCookie/}{Extensions to this
module, including a class for reading Microsoft Internet Explorer
cookies on Windows.}
\seeurl{http://www.netscape.com/newsref/std/cookie_spec.html}{The
specification of the original Netscape cookie protocol. Though this
is still the dominant protocol, the 'Netscape cookie protocol'
implemented by all the major browsers (and \module{cookielib}) only
bears a passing resemblance to the one sketched out in
\code{cookie_spec.html}.}
\seerfc{2109}{HTTP State Management Mechanism}{Obsoleted by RFC 2965.
Uses \mailheader{Set-Cookie} with version=1.}
\seerfc{2965}{HTTP State Management Mechanism}{The Netscape protocol
with the bugs fixed. Uses \mailheader{Set-Cookie2} in place of
\mailheader{Set-Cookie}. Not widely used.}
\seeurl{http://kristol.org/cookie/errata.html}{Unfinished errata to
RFC 2965.}
\seerfc{2964}{Use of HTTP State Management}{}
\end{seealso}
\subsection{CookieJar and FileCookieJar Objects \label{cookie-jar-objects}}
\class{CookieJar} objects support the iterator protocol.
\class{CookieJar} has the following methods:
\begin{methoddesc}[CookieJar]{add_cookie_header}{request}
Add correct \mailheader{Cookie} header to \var{request}.
If the CookiePolicy allows (ie. the \class{CookiePolicy} instance's
\member{rfc2965} and \member{hide_cookie2} attributes are true and
false respectively), the \mailheader{Cookie2} header is also added
when appropriate.
The \var{request} object (usually a \class{urllib2.Request} instance)
must support the methods \method{get_full_url()}, \method{get_host()},
\method{get_type()}, \method{unverifiable()},
\method{get_origin_req_host()}, \method{has_header()},
\method{get_header()}, \method{header_items()}, and
\method{add_unredirected_header()},as documented by \module{urllib2}.
\end{methoddesc}
\begin{methoddesc}[CookieJar]{extract_cookies}{response, request}
Extract cookies from HTTP \var{response} and store them in the
\class{CookieJar}, where allowed by policy.
The \class{CookieJar} will look for allowable \mailheader{Set-Cookie}
and \mailheader{Set-Cookie2} headers in the \var{response} argument,
and store cookies as appropriate (subject to the
\method{CookiePolicy.set_ok()} method's approval).
The \var{response} object (usually the result of a call to
\method{urllib2.urlopen()}, or similar) should support an
\method{info()} method, which returns an object with a
\method{getallmatchingheaders()} method (usually a
\class{mimetools.Message} instance).
The \var{request} object (usually a \class{urllib2.Request} instance)
must support the methods \method{get_full_url()}, \method{get_host()},
\method{unverifiable()}, and \method{get_origin_req_host()}, as
documented by \module{urllib2}. The request is used to set default
values for cookie-attributes as well as for checking that the cookie
is allowed to be set.
\end{methoddesc}
\begin{methoddesc}[CookieJar]{set_policy}{policy}
Set the \class{CookiePolicy} instance to be used.
\end{methoddesc}
\begin{methoddesc}[CookieJar]{make_cookies}{response, request}
Return sequence of \class{Cookie} objects extracted from
\var{response} object.
See the documentation for \method{extract_cookies} for the interfaces
required of the \var{response} and \var{request} arguments.
\end{methoddesc}
\begin{methoddesc}[CookieJar]{set_cookie_if_ok}{cookie, request}
Set a \class{Cookie} if policy says it's OK to do so.
\end{methoddesc}
\begin{methoddesc}[CookieJar]{set_cookie}{cookie}
Set a \class{Cookie}, without checking with policy to see whether or
not it should be set.
\end{methoddesc}
\begin{methoddesc}[CookieJar]{clear}{\optional{domain\optional{,
path\optional{, name}}}}
Clear some cookies.
If invoked without arguments, clear all cookies. If given a single
argument, only cookies belonging to that \var{domain} will be removed.
If given two arguments, cookies belonging to the specified
\var{domain} and URL \var{path} are removed. If given three
arguments, then the cookie with the specified \var{domain}, \var{path}
and \var{name} is removed.
Raises \exception{KeyError} if no matching cookie exists.
\end{methoddesc}
\begin{methoddesc}[CookieJar]{clear_session_cookies}{}
Discard all session cookies.
Discards all contained cookies that have a true \member{discard}
attribute (usually because they had either no \code{max-age} or
\code{expires} cookie-attribute, or an explicit \code{discard}
cookie-attribute). For interactive browsers, the end of a session
usually corresponds to closing the browser window.
Note that the \method{save()} method won't save session cookies
anyway, unless you ask otherwise by passing a true
\var{ignore_discard} argument.
\end{methoddesc}
\class{FileCookieJar} implements the following additional methods:
\begin{methoddesc}[FileCookieJar]{save}{filename=\constant{None},
ignore_discard=\constant{False}, ignore_expires=\constant{False}}
Save cookies to a file.
This base class raises \class{NotImplementedError}. Subclasses may
leave this method unimplemented.
\var{filename} is the name of file in which to save cookies. If
\var{filename} is not specified, \member{self.filename} is used (whose
default is the value passed to the constructor, if any); if
\member{self.filename} is \constant{None}, \exception{ValueError} is
raised.
\var{ignore_discard}: save even cookies set to be discarded.
\var{ignore_expires}: save even cookies that have expired
The file is overwritten if it already exists, thus wiping all the
cookies it contains. Saved cookies can be restored later using the
\method{load()} or \method{revert()} methods.
\end{methoddesc}
\begin{methoddesc}[FileCookieJar]{load}{filename=\constant{None},
ignore_discard=\constant{False}, ignore_expires=\constant{False}}
Load cookies from a file.
Old cookies are kept unless overwritten by newly loaded ones.
Arguments are as for \method{save()}.
The named file must be in the format understood by the class, or
\exception{LoadError} will be raised.
\end{methoddesc}
\begin{methoddesc}[FileCookieJar]{revert}{filename=\constant{None},
ignore_discard=\constant{False}, ignore_expires=\constant{False}}
Clear all cookies and reload cookies from a saved file.
Raises \exception{cookielib.LoadError} or \exception{IOError} if
reversion is not successful; the object's state will not be altered if
this happens.
\end{methoddesc}
\class{FileCookieJar} instances have the following public attributes:
\begin{memberdesc}{filename}
Filename of default file in which to keep cookies.
\end{memberdesc}
\begin{memberdesc}{delayload}
If true, load cookies lazily from disk. This is only a hint, since
this only affects performance, not behaviour (unless the cookies on
disk are changing). A \class{CookieJar} object may ignore it. None
of the \class{FileCookieJar} classes included in the standard library
lazily loads cookies.
\end{memberdesc}
\subsection{FileCookieJar subclasses and co-operation with web browsers
\label{file-cookie-jar-classes}}
The following \class{CookieJar} subclasses are provided for reading
and writing . Further \class{CookieJar} subclasses, including one
that reads Microsoft Internet Explorer cookies, are available at
\url{http://wwwsearch.sf.net/ClientCookie/}.
\begin{classdesc}{MozillaCookieJar}{filename, delayload=\constant{None},
policy=\constant{None}}
A \class{FileCookieJar} that can load from and save cookies to disk in
the Mozilla \code{cookies.txt} file format (which is also used by the
lynx and Netscape browsers). \note{This loses information about RFC
2965 cookies, and also about newer or non-standard cookie-attributes
such as \code{port}.}
\warning{Back up your cookies before saving if you have cookies whose
loss / corruption would be inconvenient (there are some subtleties
which may lead to slight changes in the file over a load / save
round-trip).}
Also note that cookies saved while Mozilla is running will get
clobbered by Mozilla.
\end{classdesc}
\begin{classdesc}{LWPCookieJar}{filename, delayload=\constant{None},
policy=\constant{None}}
A \class{FileCookieJar} that can load from and save cookies to disk in
format compatible with the libwww-perl library's \code{Set-Cookie3}
file format. This is convenient if you want to store cookies in a
human-readable file.
\end{classdesc}
\subsection{CookiePolicy Objects \label{cookie-policy-objects}}
Objects implementing the \class{CookiePolicy} interface have the
following methods:
\begin{methoddesc}[CookiePolicy]{set_ok}{cookie, request}
Return boolean value indicating whether cookie should be accepted from server.
\var{cookie} is a \class{cookielib.Cookie} instance. \var{request} is
an object implementing the interface defined by the documentation for
\method{CookieJar.extract_cookies()}.
\end{methoddesc}
\begin{methoddesc}[CookiePolicy]{return_ok}{cookie, request}
Return boolean value indicating whether cookie should be returned to server.
\var{cookie} is a \class{cookielib.Cookie} instance. \var{request} is
an object implementing the interface defined by the documentation for
\method{CookieJar.add_cookie_header()}.
\end{methoddesc}
\begin{methoddesc}[CookiePolicy]{domain_return_ok}{domain, request}
Return false if cookies should not be returned, given cookie domain.
This method is an optimization. It removes the need for checking
every cookie with a particular domain (which might involve reading
many files). The default implementations of
\method{domain_return_ok()} and \method{path_return_ok()}
(\samp{return True}) leave all the work to \method{return_ok()}.
If \method{domain_return_ok()} returns true for the cookie domain,
\method{path_return_ok()} is called for the cookie path. Otherwise,
\method{path_return_ok()} and \method{return_ok()} are never called
for that cookie domain. If \method{path_return_ok()} returns true,
\method{return_ok()} is called with the \class{Cookie} object itself
for a full check. Otherwise, \method{return_ok()} is never called for
that cookie path.
Note that \method{domain_return_ok()} is called for every
\emph{cookie} domain, not just for the \emph{request} domain. For
example, the function might be called with both \code{".example.com"}
and \code{"www.example.com"} if the request domain is
\code{"www.example.com"}. The same goes for
\method{path_return_ok()}.
The \var{request} argument is as documented for \method{return_ok()}.
\end{methoddesc}
\begin{methoddesc}[CookiePolicy]{path_return_ok}{path, request}
Return false if cookies should not be returned, given cookie path.
See the documentation for \method{domain_return_ok()}.
\end{methoddesc}
In addition to implementing the methods above, implementations of the
\class{CookiePolicy} interface must also supply the following
attributes, indicating which protocols should be used, and how. All
of these attributes may be assigned to.
\begin{memberdesc}{netscape}
Implement netscape protocol.
\end{memberdesc}
\begin{memberdesc}{rfc2965}
Implement RFC 2965 protocol.
\end{memberdesc}
\begin{memberdesc}{hide_cookie2}
Don't add Cookie2 header to requests (the presence of this header
indicates to the server that we understand RFC 2965 cookies).
\end{memberdesc}
The most useful way to define a \class{CookiePolicy} class is by
subclassing from \class{DefaultCookiePolicy} and overriding some or
all of the methods above. \class{CookiePolicy} itself may be used as
a 'null policy' to allow setting and receiving any and all cookies.
\subsection{DefaultCookiePolicy Objects \label{default-cookie-policy-objects}}
Implements the standard rules for accepting and returning cookies.
Both RFC 2965 and Netscape cookies are covered. RFC 2965 handling is
switched off by default.
The easiest way to provide your own policy is to override this class
and call its methods in your overriden implementations before adding
your own additional checks:
\begin{verbatim}
import cookielib
class MyCookiePolicy(cookielib.DefaultCookiePolicy):
def set_ok(self, cookie, request):
if not cookielib.DefaultCookiePolicy.set_ok(self, cookie, request):
return False
if i_dont_want_to_store_this_cookie(cookie):
return False
return True
\end{verbatim}
In addition to the features required to implement the
\class{CookiePolicy} interface, this class allows you to block and
allow domains from setting and receiving cookies. There are also some
strictness switches that allow you to tighten up the rather loose
Netscape protocol rules a little bit (at the cost of blocking some
benign cookies).
A domain blacklist and whitelist is provided (both off by default).
Only domains not in the blacklist and present in the whitelist (if the
whitelist is active) participate in cookie setting and returning. Use
the \var{blocked_domains} constructor argument, and
\method{blocked_domains()} and \method{set_blocked_domains()} methods
(and the corresponding argument and methods for
\var{allowed_domains}). If you set a whitelist, you can turn it off
again by setting it to \constant{None}.
Domains in block or allow lists that do not start with a dot must be
equal. For example, \code{"example.com"} matches a blacklist entry of
\code{"example.com"}, but \code{"www.example.com"} does not. Domains
that do start with a dot are matched by more specific domains too.
For example, both \code{"www.example.com"} and
\code{"www.coyote.example.com"} match \code{".example.com"} (but
\code{"example.com"} itself does not). IP addresses are an exception,
and must match exactly. For example, if blocked_domains contains
\code{"192.168.1.2"} and \code{".168.1.2"}, 192.168.1.2 is blocked,
but 193.168.1.2 is not.
\class{DefaultCookiePolicy} implements the following additional
methods:
\begin{methoddesc}[DefaultCookiePolicy]{blocked_domains}{}
Return the sequence of blocked domains (as a tuple).
\end{methoddesc}
\begin{methoddesc}[DefaultCookiePolicy]{set_blocked_domains}
{blocked_domains}
Set the sequence of blocked domains.
\end{methoddesc}
\begin{methoddesc}[DefaultCookiePolicy]{is_blocked}{domain}
Return whether \var{domain} is on the blacklist for setting or
receiving cookies.
\end{methoddesc}
\begin{methoddesc}[DefaultCookiePolicy]{allowed_domains}{}
Return \constant{None}, or the sequence of allowed domains (as a tuple).
\end{methoddesc}
\begin{methoddesc}[DefaultCookiePolicy]{set_allowed_domains}
{allowed_domains}
Set the sequence of allowed domains, or \constant{None}.
\end{methoddesc}
\begin{methoddesc}[DefaultCookiePolicy]{is_not_allowed}{domain}
Return whether \var{domain} is not on the whitelist for setting or
receiving cookies.
\end{methoddesc}
\class{DefaultCookiePolicy} instances have the following attributes,
which are all initialised from the constructor arguments of the same
name, and which may all be assigned to.
General strictness switches:
\begin{memberdesc}{strict_domain}
Don't allow sites to set two-component domains with country-code
top-level domains like \code{.co.uk}, \code{.gov.uk},
\code{.co.nz}.etc. This is far from perfect and isn't guaranteed to
work!
\end{memberdesc}
RFC 2965 protocol strictness switches:
\begin{memberdesc}{strict_rfc2965_unverifiable}
Follow RFC 2965 rules on unverifiable transactions (usually, an
unverifiable transaction is one resulting from a redirect or a request
for an image hosted on another site). If this is false, cookies are
\emph{never} blocked on the basis of verifiability
\end{memberdesc}
Netscape protocol strictness switches:
\begin{memberdesc}{strict_ns_unverifiable}
apply RFC 2965 rules on unverifiable transactions even to Netscape
cookies
\end{memberdesc}
\begin{memberdesc}{strict_ns_domain}
Flags indicating how strict to be with domain-matching rules for
Netscape cookies. See below for acceptable values.
\end{memberdesc}
\begin{memberdesc}{strict_ns_set_initial_dollar}
Ignore cookies in Set-Cookie: headers that have names starting with
\code{'\$'}.
\end{memberdesc}
\begin{memberdesc}{strict_ns_set_path}
Don't allow setting cookies whose path doesn't path-match request URI.
\end{memberdesc}
\member{strict_ns_domain} is a collection of flags. Its value is
constructed by or-ing together (for example,
\code{DomainStrictNoDots|DomainStrictNonDomain} means both flags are
set).
\begin{memberdesc}{DomainStrictNoDots}
When setting cookies, the 'host prefix' must not contain a dot
(eg. \code{www.foo.bar.com} can't set a cookie for \code{.bar.com},
because \code{www.foo} contains a dot).
\end{memberdesc}
\begin{memberdesc}{DomainStrictNonDomain}
Cookies that did not explicitly specify a \code{domain}
cookie-attribute can only be returned to a domain that string-compares
equal to the domain that set the cookie (eg. \code{spam.example.com}
won't be returned cookies from \code{example.com} that had no
\code{domain} cookie-attribute).
\end{memberdesc}
\begin{memberdesc}{DomainRFC2965Match}
When setting cookies, require a full RFC 2965 domain-match.
\end{memberdesc}
The following attributes are provided for convenience, and are the
most useful combinations of the above flags:
\begin{memberdesc}{DomainLiberal}
Equivalent to 0 (ie. all of the above Netscape domain strictness flags
switched off).
\end{memberdesc}
\begin{memberdesc}{DomainStrict}
Equivalent to \code{DomainStrictNoDots|DomainStrictNonDomain}.
\end{memberdesc}
\subsection{Cookie Objects \label{cookie-jar-objects}}
\class{Cookie} instances have Python attributes roughly corresponding
to the standard cookie-attributes specified in the various cookie
standards. The correspondence is not one-to-one, because there are
complicated rules for assigning default values, and because the
\code{max-age} and \code{expires} cookie-attributes contain equivalent
information.
Assignment to these attributes should not be necessary other than in
rare circumstances in a \class{CookiePolicy} method. The class does
not enforce internal consistency, so you should know what you're
doing if you do that.
\begin{memberdesc}[Cookie]{version}
Integer or \constant{None}. Netscape cookies have version 0. RFC
2965 and RFC 2109 cookies have version 1.
\end{memberdesc}
\begin{memberdesc}[Cookie]{name}
Cookie name (a string), or \constant{None}.
\end{memberdesc}
\begin{memberdesc}[Cookie]{value}
Cookie value (a string).
\end{memberdesc}
\begin{memberdesc}[Cookie]{port}
String representing a port or a set of ports (eg. '80', or '80,8080'),
or \constant{None}.
\end{memberdesc}
\begin{memberdesc}[Cookie]{path}
Cookie path (a string, eg. '/acme/rocket_launchers').
\end{memberdesc}
\begin{memberdesc}[Cookie]{secure}
True if cookie should only be returned over a secure connection.
\end{memberdesc}
\begin{memberdesc}[Cookie]{expires}
Integer expiry date in seconds since epoch, or \constant{None}. See
also the \method{is_expired()} method.
\end{memberdesc}
\begin{memberdesc}[Cookie]{discard}
True if this is a session cookie.
\end{memberdesc}
\begin{memberdesc}[Cookie]{comment}
String comment from the server explaining the function of this cookie,
or \constant{None}.
\end{memberdesc}
\begin{memberdesc}[Cookie]{comment_url}
URL linking to a comment from the server explaining the function of
this cookie, or \constant{None}.
\end{memberdesc}
\begin{memberdesc}[Cookie]{port_specified}
True if a port or set of ports was explicitly specified by the server
(in the \mailheader{Set-Cookie} / \mailheader{Set-Cookie2} header).
\end{memberdesc}
\begin{memberdesc}[Cookie]{domain_specified}
True if a domain was explicitly specified by the server.
\end{memberdesc}
\begin{memberdesc}[Cookie]{domain_initial_dot}
True if the domain explicitly specified by the server began with a
dot ('.').
\end{memberdesc}
Cookies may have additional non-standard cookie-attributes. These may
be accessed using the following methods:
\begin{methoddesc}[Cookie]{has_nonstandard_attr}{name}
Return true if cookie has the named cookie-attribute.
\end{methoddesc}
\begin{methoddesc}[Cookie]{get_nonstandard_attr}{name, default=\constant{None}}
If cookie has the named cookie-attribute, return its value.
Otherwise, return \var{default}.
\end{methoddesc}
\begin{methoddesc}[Cookie]{set_nonstandard_attr}{name, value}
Set the value of the named cookie-attribute.
\end{methoddesc}
The \class{Cookie} class also defines the following method:
\begin{methoddesc}[Cookie]{is_expired}{\optional{now=\constant{None}}}
True if cookie has passed the time at which the server requested it
should expire. If \var{now} is given (in seconds since the epoch),
return whether the cookie has expired at the specified time.
\end{methoddesc}
\subsection{Examples \label{cookielib-examples}}
The first example shows the most common usage of \module{cookielib}:
\begin{verbatim}
import cookielib, urllib2
cj = cookielib.CookieJar()
opener = urllib2.build_opener(urllib2.HTTPCookieProcessor(cj))
r = opener.open("http://example.com/")
\end{verbatim}
This example illustrates how to open a URL using your Netscape,
Mozilla, or lynx cookies (assumes \UNIX{} convention for location of
the cookies file):
\begin{verbatim}
import os, cookielib, urllib2
cj = cookielib.MozillaCookieJar()
cj.load(os.path.join(os.environ["HOME"], "/.netscape/cookies.txt"))
opener = urllib2.build_opener(urllib2.HTTPCookieProcessor(cj))
r = opener.open("http://example.com/")
\end{verbatim}
The next example illustrates the use of \class{DefaultCookiePolicy}.
Turn on RFC 2965 cookies, be more strict about domains when setting
and returning Netscape cookies, and block some domains from setting
cookies or having them returned:
\begin{verbatim}
import urllib2
from cookielib import CookieJar, DefaultCookiePolicy as Policy
policy = Policy(rfc2965=True, strict_ns_domain=Policy.DomainStrict,
blocked_domains=["ads.net", ".ads.net"])
cj = CookieJar(policy)
opener = urllib2.build_opener(urllib2.HTTPCookieProcessor(cj))
r = opener.open("http://example.com/")
\end{verbatim}

161
Doc/lib/liburllib2.tex
View File

@ -10,14 +10,13 @@
The \module{urllib2} module defines functions and classes which help
in opening URLs (mostly HTTP) in a complex world --- basic and digest
authentication, redirections and more.
authentication, redirections, cookies and more.
The \module{urllib2} module defines the following functions:
\begin{funcdesc}{urlopen}{url\optional{, data}}
Open the URL \var{url}, which can be either a string or a \class{Request}
object (currently the code checks that it really is a \class{Request}
instance, or an instance of a subclass of \class{Request}).
object.
\var{data} should be a string, which specifies additional data to
send to the server. In HTTP requests, which are the only ones that
@ -87,13 +86,32 @@ Gopher handler.
The following classes are provided:
\begin{classdesc}{Request}{url\optional{, data\optional{, headers}}}
\begin{classdesc}{Request}{url\optional{, data}\optional{, headers}
\optional{, origin_req_host}\optional{, unverifiable}}
This class is an abstraction of a URL request.
\var{url} should be a string which is a valid URL. For a description
of \var{data} see the \method{add_data()} description.
\var{headers} should be a dictionary, and will be treated as if
\method{add_header()} was called with each key and value as arguments.
The final two arguments are only of interest for correct handling of
third-party HTTP cookies:
\var{origin_req_host} should be the request-host of the origin
transaction, as defined by \rfc{2965}. It defaults to
\code{cookielib.request_host(self)}. This is the host name or IP
address of the original request that was initiated by the user. For
example, if the request is for an image in an HTML document, this
should be the request-host of the request for the page containing the
image.
\var{unverifiable} should indicate whether the request is
unverifiable, as defined by RFC 2965. It defaults to False. An
unverifiable request is one whose URL the user did not have the option
to approve. For example, if the request is for an image in an HTML
document, and the user had no option to approve the automatic fetching
of the image, this should be true.
\end{classdesc}
\begin{classdesc}{OpenerDirector}{}
@ -116,6 +134,10 @@ responses are turned into \exception{HTTPError} exceptions.
A class to handle redirections.
\end{classdesc}
\begin{classdesc}{HTTPCookieProcessor}{\optional{cookiejar}}
A class to handle HTTP Cookies.
\end{classdesc}
\begin{classdesc}{ProxyHandler}{\optional{proxies}}
Cause requests to go through a proxy.
If \var{proxies} is given, it must be a dictionary mapping
@ -217,10 +239,10 @@ The following methods describe all of \class{Request}'s public interface,
and so all must be overridden in subclasses.
\begin{methoddesc}[Request]{add_data}{data}
Set the \class{Request} data to \var{data}. This is ignored
by all handlers except HTTP handlers --- and there it should be an
\mimetype{application/x-www-form-encoded} buffer, and will change the
request to be \code{POST} rather than \code{GET}.
Set the \class{Request} data to \var{data}. This is ignored by all
handlers except HTTP handlers --- and there it should be a byte
string, and will change the request to be \code{POST} rather than
\code{GET}.
\end{methoddesc}
\begin{methoddesc}[Request]{get_method}{}
@ -282,6 +304,17 @@ and \var{type} will replace those of the instance, and the instance's
selector will be the original URL given in the constructor.
\end{methoddesc}
\begin{methoddesc}[Request]{get_origin_req_host}{}
Return the request-host of the origin transaction, as defined by
\rfc{2965}. See the documentation for the \class{Request}
constructor.
\end{methoddesc}
\begin{methoddesc}[Request]{is_unverifiable}{}
Return whether the request is unverifiable, as defined by RFC 2965.
See the documentation for the \class{Request} constructor.
\end{methoddesc}
\subsection{OpenerDirector Objects \label{opener-director-objects}}
@ -289,14 +322,18 @@ selector will be the original URL given in the constructor.
\begin{methoddesc}[OpenerDirector]{add_handler}{handler}
\var{handler} should be an instance of \class{BaseHandler}. The
following methods are searched, and added to the possible chains.
following methods are searched, and added to the possible chains (note
that HTTP errors are a special case).
\begin{itemize}
\item \method{\var{protocol}_open()} ---
signal that the handler knows how to open \var{protocol} URLs.
\item \method{\var{protocol}_error_\var{type}()} ---
signal that the handler knows how to handle \var{type} errors from
\var{protocol}.
\item \method{http_error_\var{type}()} ---
signal that the handler knows how to handle HTTP errors with HTTP
error code \var{type}.
\item \method{\var{protocol}_error()} ---
signal that the handler knows how to handle errors from
(non-\code{http}) \var{protocol}.
\item \method{\var{protocol}_request()} ---
signal that the handler knows how to pre-process \var{protocol}
requests.
@ -306,26 +343,17 @@ following methods are searched, and added to the possible chains.
\end{itemize}
\end{methoddesc}
\begin{methoddesc}[OpenerDirector]{close}{}
Explicitly break cycles, and delete all the handlers.
Because the \class{OpenerDirector} needs to know the registered handlers,
and a handler needs to know who the \class{OpenerDirector} who called
it is, there is a reference cycle. Even though recent versions of Python
have cycle-collection, it is sometimes preferable to explicitly break
the cycles.
\end{methoddesc}
\begin{methoddesc}[OpenerDirector]{open}{url\optional{, data}}
Open the given \var{url} (which can be a request object or a string),
optionally passing the given \var{data}.
Arguments, return values and exceptions raised are the same as those
of \function{urlopen()} (which simply calls the \method{open()} method
on the default installed \class{OpenerDirector}).
on the currently installed global \class{OpenerDirector}).
\end{methoddesc}
\begin{methoddesc}[OpenerDirector]{error}{proto\optional{,
arg\optional{, \moreargs}}}
Handle an error in a given protocol. This will call the registered
Handle an error of the given protocol. This will call the registered
error handlers for the given protocol with the given arguments (which
are protocol specific). The HTTP protocol is a special case which
uses the HTTP response code to determine the specific error handler;
@ -335,6 +363,45 @@ Return values and exceptions raised are the same as those
of \function{urlopen()}.
\end{methoddesc}
OpenerDirector objects open URLs in three stages:
\begin{enumerate}
\item Every handler with a method named like
\method{\var{protocol}_request()} has that method called to
pre-process the request.
The order in which these methods are called is determined by
sorting the handler instances by the \member{.processor_order}
attribute.
\item Handlers with a method named like
\method{\var{protocol}_open()} are called to handle the request.
This stage ends when a handler either returns a
non-\constant{None} value (ie. a response), or raises an exception
(usually URLError). Exceptions are allowed to propagate.
In fact, the above algorithm is first tried for methods named
\method{default_open}. If all such methods return
\constant{None}, the algorithm is repeated for methods named like
\method{\var{protocol}_open()}. If all such methods return
\constant{None}, the algorithm is repeated for methods named
\method{unknown_open()}.
Note that the implementation of these methods may involve calls of
the parent \class{OpenerDirector} instance's \method{.open()} and
\method{.error()} methods.
The order in which these methods are called is determined by
sorting the handler instances.
\item Every handler with a method named like
\method{\var{protocol}_response()} has that method called to
post-process the response.
The order in which these methods are called is determined by
sorting the handler instances by the \member{.processor_order}
attribute.
\end{enumerate}
\subsection{BaseHandler Objects \label{base-handler-objects}}
@ -351,7 +418,11 @@ Remove any parents.
\end{methoddesc}
The following members and methods should only be used by classes
derived from \class{BaseHandler}:
derived from \class{BaseHandler}. \note{The convention has been
adopted that subclasses defining \method{\var{protocol}_request()} or
\method{\var{protocol}_response()} methods are named
\class{*Processor}; all others are named \class{*Handler}.}
\begin{memberdesc}[BaseHandler]{parent}
A valid \class{OpenerDirector}, which can be used to open using a
@ -423,6 +494,29 @@ Arguments, return values and exceptions raised should be the same as
for \method{http_error_default()}.
\end{methoddesc}
\begin{methoddescni}[BaseHandler]{\var{protocol}_request}{req}
This method is \emph{not} defined in \class{BaseHandler}, but
subclasses should define it if they want to pre-process requests of
the given protocol.
This method, if defined, will be called by the parent
\class{OpenerDirector}. \var{req} will be a \class{Request} object.
The return value should be a \class{Request} object.
\end{methoddescni}
\begin{methoddescni}[BaseHandler]{\var{protocol}_response}{req, response}
This method is \emph{not} defined in \class{BaseHandler}, but
subclasses should define it if they want to post-process responses of
the given protocol.
This method, if defined, will be called by the parent
\class{OpenerDirector}. \var{req} will be a \class{Request} object.
\var{response} will be an object implementing the same interface as
the return value of \function{urlopen()}. The return value should
implement the same interface as the return value of
\function{urlopen()}.
\end{methoddescni}
\subsection{HTTPRedirectHandler Objects \label{http-redirect-handler}}
\note{Some HTTP redirections require action from this module's client
@ -434,12 +528,12 @@ for \method{http_error_default()}.
fp, code, msg, hdrs}
Return a \class{Request} or \code{None} in response to a redirect.
This is called by the default implementations of the
\method{http_error_30*()} methods when a redirection is received
from the server. If a redirection should take place, return a new
\method{http_error_30*()} methods when a redirection is received from
the server. If a redirection should take place, return a new
\class{Request} to allow \method{http_error_30*()} to perform the
redirect. Otherwise, raise \exception{HTTPError} if no other
\class{Handler} should try to handle this URL, or return \code{None}
if you can't but another \class{Handler} might.
redirect. Otherwise, raise \exception{HTTPError} if no other handler
should try to handle this URL, or return \code{None} if you can't but
another handler might.
\begin{notice}
The default implementation of this method does not strictly
@ -478,6 +572,15 @@ The same as \method{http_error_301()}, but called for the
\end{methoddesc}
\subsection{HTTPCookieProcessor Objects \label{http-cookie-processor}}
\class{HTTPCookieProcessor} instances have one attribute:
\begin{memberdesc}{cookiejar}
The \class{cookielib.CookieJar} in which cookies are stored.
\end{memberdesc}
\subsection{ProxyHandler Objects \label{proxy-handler}}
\begin{methoddescni}[ProxyHandler]{\var{protocol}_open}{request}

19
Doc/whatsnew/whatsnew24.tex
View File

@ -569,6 +569,25 @@ For example:
%======================================================================
% whole new modules get described in \subsections here
\subsection{cookielib}
The \module{cookielib} library supports client-side handling for HTTP
cookies, just as the \module{Cookie} provides server-side cookie
support in CGI scripts. This library manages cookies in a way similar
to web browsers. Cookies are stored in cookie jars; the library
transparently stores cookies offered by the web server in the cookie
jar, and fetches the cookie from the jar when connecting to the
server. Similar to web browsers, policy objects control whether
cookies are accepted or not.
In order to store cookies across sessions, two implementations of
cookie jars are provided: one that stores cookies in the Netscape
format, so applications can use the Mozilla or Lynx cookie jars, and
one that stores cookies in the same format as the Perl libwww libary.
\module{urllib2} has been changed to interact with \module{cookielib}:
\class{HTTPCookieProcessor} manages a cookie jar that is used when
accessing URLs.
% ======================================================================
\section{Build and C API Changes}

171
Lib/_LWPCookieJar.py Normal file
View File

@ -0,0 +1,171 @@
"""Load / save to libwww-perl (LWP) format files.
Actually, the format is slightly extended from that used by LWP's
(libwww-perl's) HTTP::Cookies, to avoid losing some RFC 2965 information
not recorded by LWP.
It uses the version string "2.0", though really there isn't an LWP Cookies
2.0 format. This indicates that there is extra information in here
(domain_dot and # port_spec) while still being compatible with
libwww-perl, I hope.
"""
import time, re, logging
from cookielib import reraise_unmasked_exceptions, FileCookieJar, Cookie, \
MISSING_FILENAME_TEXT, join_header_words, split_header_words, \
iso2time, time2isoz
def lwp_cookie_str(cookie):
"""Return string representation of Cookie in an the LWP cookie file format.
Actually, the format is extended a bit -- see module docstring.
"""
h = [(cookie.name, cookie.value),
("path", cookie.path),
("domain", cookie.domain)]
if cookie.port is not None: h.append(("port", cookie.port))
if cookie.path_specified: h.append(("path_spec", None))
if cookie.port_specified: h.append(("port_spec", None))
if cookie.domain_initial_dot: h.append(("domain_dot", None))
if cookie.secure: h.append(("secure", None))
if cookie.expires: h.append(("expires",
time2isoz(float(cookie.expires))))
if cookie.discard: h.append(("discard", None))
if cookie.comment: h.append(("comment", cookie.comment))
if cookie.comment_url: h.append(("commenturl", cookie.comment_url))
keys = cookie._rest.keys()
keys.sort()
for k in keys:
h.append((k, str(cookie._rest[k])))
h.append(("version", str(cookie.version)))
return join_header_words([h])
class LWPCookieJar(FileCookieJar):
"""
The LWPCookieJar saves a sequence of"Set-Cookie3" lines.
"Set-Cookie3" is the format used by the libwww-perl libary, not known
to be compatible with any browser, but which is easy to read and
doesn't lose information about RFC 2965 cookies.
Additional methods
as_lwp_str(ignore_discard=True, ignore_expired=True)
"""
def as_lwp_str(self, ignore_discard=True, ignore_expires=True):
"""Return cookies as a string of "\n"-separated "Set-Cookie3" headers.
ignore_discard and ignore_expires: see docstring for FileCookieJar.save
"""
now = time.time()
r = []
for cookie in self:
if not ignore_discard and cookie.discard:
continue
if not ignore_expires and cookie.is_expired(now):
continue
r.append("Set-Cookie3: %s" % lwp_cookie_str(cookie))
return "\n".join(r+[""])
def save(self, filename=None, ignore_discard=False, ignore_expires=False):
if filename is None:
if self.filename is not None: filename = self.filename
else: raise ValueError(MISSING_FILENAME_TEXT)
f = open(filename, "w")
try:
# There really isn't an LWP Cookies 2.0 format, but this indicates
# that there is extra information in here (domain_dot and
# port_spec) while still being compatible with libwww-perl, I hope.
f.write("#LWP-Cookies-2.0\n")
f.write(self.as_lwp_str(ignore_discard, ignore_expires))
finally:
f.close()
def _really_load(self, f, filename, ignore_discard, ignore_expires):
magic = f.readline()
if not re.search(self.magic_re, magic):
msg = "%s does not seem to contain cookies" % filename
raise IOError(msg)
now = time.time()
header = "Set-Cookie3:"
boolean_attrs = ("port_spec", "path_spec", "domain_dot",
"secure", "discard")
value_attrs = ("version",
"port", "path", "domain",
"expires",
"comment", "commenturl")
try:
while 1:
line = f.readline()
if line == "": break
if not line.startswith(header):
continue
line = line[len(header):].strip()
for data in split_header_words([line]):
name, value = data[0]
# name and value are an exception here, since a plain "foo"
# (with no "=", unlike "bar=foo") means a cookie with no
# name and value "foo". With all other cookie-attributes,
# the situation is reversed: "foo" means an attribute named
# "foo" with no value!
if value is None:
name, value = value, name
standard = {}
rest = {}
for k in boolean_attrs:
standard[k] = False
for k, v in data[1:]:
if k is not None:
lc = k.lower()
else:
lc = None
# don't lose case distinction for unknown fields
if (lc in value_attrs) or (lc in boolean_attrs):
k = lc
if k in boolean_attrs:
if v is None: v = True
standard[k] = v
elif k in value_attrs:
standard[k] = v
else:
rest[k] = v
h = standard.get
expires = h("expires")
discard = h("discard")
if expires is not None:
expires = iso2time(expires)
if expires is None:
discard = True
domain = h("domain")
domain_specified = domain.startswith(".")
c = Cookie(h("version"), name, value,
h("port"), h("port_spec"),
domain, domain_specified, h("domain_dot"),
h("path"), h("path_spec"),
h("secure"),
expires,
discard,
h("comment"),
h("commenturl"),
rest)
if not ignore_discard and c.discard:
continue
if not ignore_expires and c.is_expired(now):
continue
self.set_cookie(c)
except:
reraise_unmasked_exceptions((IOError,))
raise IOError("invalid Set-Cookie3 format file %s" % filename)

145
Lib/_MozillaCookieJar.py Normal file
View File

@ -0,0 +1,145 @@
"""Mozilla / Netscape cookie loading / saving."""
import re, time, logging
from cookielib import reraise_unmasked_exceptions, FileCookieJar, Cookie, \
MISSING_FILENAME_TEXT
class MozillaCookieJar(FileCookieJar):
"""
WARNING: you may want to backup your browser's cookies file if you use
this class to save cookies. I *think* it works, but there have been
bugs in the past!
This class differs from CookieJar only in the format it uses to save and
load cookies to and from a file. This class uses the Mozilla/Netscape
`cookies.txt' format. lynx uses this file format, too.
Don't expect cookies saved while the browser is running to be noticed by
the browser (in fact, Mozilla on unix will overwrite your saved cookies if
you change them on disk while it's running; on Windows, you probably can't
save at all while the browser is running).
Note that the Mozilla/Netscape format will downgrade RFC2965 cookies to
Netscape cookies on saving.
In particular, the cookie version and port number information is lost,
together with information about whether or not Path, Port and Discard were
specified by the Set-Cookie2 (or Set-Cookie) header, and whether or not the
domain as set in the HTTP header started with a dot (yes, I'm aware some
domains in Netscape files start with a dot and some don't -- trust me, you
really don't want to know any more about this).
Note that though Mozilla and Netscape use the same format, they use
slightly different headers. The class saves cookies using the Netscape
header by default (Mozilla can cope with that).
"""
magic_re = "#( Netscape)? HTTP Cookie File"
header = """\
# Netscape HTTP Cookie File
# http://www.netscape.com/newsref/std/cookie_spec.html
# This is a generated file! Do not edit.
"""
def _really_load(self, f, filename, ignore_discard, ignore_expires):
now = time.time()
magic = f.readline()
if not re.search(self.magic_re, magic):
f.close()
raise IOError(
"%s does not look like a Netscape format cookies file" %
filename)
try:
while 1:
line = f.readline()
if line == "": break
# last field may be absent, so keep any trailing tab
if line.endswith("\n"): line = line[:-1]
# skip comments and blank lines XXX what is $ for?
if (line.strip().startswith("#") or
line.strip().startswith("$") or
line.strip() == ""):
continue
domain, domain_specified, path, secure, expires, name, value = \
line.split("\t")
secure = (secure == "TRUE")
domain_specified = (domain_specified == "TRUE")
if name == "":
name = value
value = None
initial_dot = domain.startswith(".")
assert domain_specified == initial_dot
discard = False
if expires == "":
expires = None
discard = True
# assume path_specified is false
c = Cookie(0, name, value,
None, False,
domain, domain_specified, initial_dot,
path, False,
secure,
expires,
discard,
None,
None,
{})
if not ignore_discard and c.discard:
continue
if not ignore_expires and c.is_expired(now):
continue
self.set_cookie(c)
except:
reraise_unmasked_exceptions((IOError,))
raise IOError("invalid Netscape format file %s: %s" %
(filename, line))
def save(self, filename=None, ignore_discard=False, ignore_expires=False):
if filename is None:
if self.filename is not None: filename = self.filename
else: raise ValueError(MISSING_FILENAME_TEXT)
f = open(filename, "w")
try:
f.write(self.header)
now = time.time()
for cookie in self:
if not ignore_discard and cookie.discard:
continue
if not ignore_expires and cookie.is_expired(now):
continue
if cookie.secure: secure = "TRUE"
else: secure = "FALSE"
if cookie.domain.startswith("."): initial_dot = "TRUE"
else: initial_dot = "FALSE"
if cookie.expires is not None:
expires = str(cookie.expires)
else:
expires = ""
if cookie.value is None:
# cookies.txt regards 'Set-Cookie: foo' as a cookie
# with no name, whereas cookielib regards it as a
# cookie with no value.
name = ""
value = cookie.name
else:
name = cookie.name
value = cookie.value
f.write(
"\t".join([cookie.domain, initial_dot, cookie.path,
secure, expires, name, value])+
"\n")
finally:
f.close()

1756
Lib/cookielib.py Normal file
View File

File diff suppressed because it is too large Load Diff

1620
Lib/test/test_cookielib.py Normal file
View File

File diff suppressed because it is too large Load Diff

101
Lib/test/test_urllib2.py
View File

@ -54,6 +54,10 @@ class MockFile:
def readline(self, count=None): pass
def close(self): pass
class MockHeaders(dict):
def getheaders(self, name):
return self.values()
class MockResponse(StringIO.StringIO):
def __init__(self, code, msg, headers, data, url=None):
StringIO.StringIO.__init__(self, data)
@ -63,6 +67,12 @@ class MockResponse(StringIO.StringIO):
def geturl(self):
return self.url
class MockCookieJar:
def add_cookie_header(self, request):
self.ach_req = request
def extract_cookies(self, response, request):
self.ec_req, self.ec_r = request, response
class FakeMethod:
def __init__(self, meth_name, action, handle):
self.meth_name = meth_name
@ -474,7 +484,7 @@ class HandlerTests(unittest.TestCase):
for data in "", None: # POST, GET
req = Request("http://example.com/", data)
r = MockResponse(200, "OK", {}, "")
newreq = h.do_request(req)
newreq = h.do_request_(req)
if data is None: # GET
self.assert_("Content-length" not in req.unredirected_hdrs)
self.assert_("Content-type" not in req.unredirected_hdrs)
@ -491,7 +501,7 @@ class HandlerTests(unittest.TestCase):
req.add_unredirected_header("Content-type", "bar")
req.add_unredirected_header("Host", "baz")
req.add_unredirected_header("Spam", "foo")
newreq = h.do_request(req)
newreq = h.do_request_(req)
self.assertEqual(req.unredirected_hdrs["Content-length"], "foo")
self.assertEqual(req.unredirected_hdrs["Content-type"], "bar")
self.assertEqual(req.unredirected_hdrs["Host"], "baz")
@ -514,6 +524,21 @@ class HandlerTests(unittest.TestCase):
self.assertEqual(o.proto, "http") # o.error called
self.assertEqual(o.args, (req, r, 201, "Created", {}))
def test_cookies(self):
cj = MockCookieJar()
h = urllib2.HTTPCookieProcessor(cj)
o = h.parent = MockOpener()
req = Request("http://example.com/")
r = MockResponse(200, "OK", {}, "")
newreq = h.http_request(req)
self.assert_(cj.ach_req is req is newreq)
self.assertEquals(req.get_origin_req_host(), "example.com")
self.assert_(not req.is_unverifiable())
newr = h.http_response(req, r)
self.assert_(cj.ec_req is req)
self.assert_(cj.ec_r is r is newr)
def test_redirect(self):
from_url = "http://example.com/a.html"
to_url = "http://example.com/b.html"
@ -528,7 +553,8 @@ class HandlerTests(unittest.TestCase):
req.add_header("Nonsense", "viking=withhold")
req.add_unredirected_header("Spam", "spam")
try:
method(req, MockFile(), code, "Blah", {"location": to_url})
method(req, MockFile(), code, "Blah",
MockHeaders({"location": to_url}))
except urllib2.HTTPError:
# 307 in response to POST requires user OK
self.assert_(code == 307 and data is not None)
@ -544,38 +570,65 @@ class HandlerTests(unittest.TestCase):
# loop detection
req = Request(from_url)
req.origin_req_host = "example.com"
def redirect(h, req, code, url=to_url):
method = getattr(h, "http_error_%s" % code)
method(req, MockFile(), code, "Blah", {"location": url})
def redirect(h, req, url=to_url):
h.http_error_302(req, MockFile(), 302, "Blah",
MockHeaders({"location": url}))
# Note that the *original* request shares the same record of
# redirections with the sub-requests caused by the redirections.
# once
redirect(h, req, 302)
# twice: loop detected
self.assertRaises(urllib2.HTTPError, redirect, h, req, 302)
# and again
self.assertRaises(urllib2.HTTPError, redirect, h, req, 302)
# but this is a different redirect code, so OK...
redirect(h, req, 301)
self.assertRaises(urllib2.HTTPError, redirect, h, req, 301)
# order doesn't matter
redirect(h, req, 303)
redirect(h, req, 307)
self.assertRaises(urllib2.HTTPError, redirect, h, req, 303)
# detect endless non-repeating chain of redirects
req = Request(from_url)
req.origin_req_host = "example.com"
# detect infinite loop redirect of a URL to itself
req = Request(from_url, origin_req_host="example.com")
count = 0
try:
while 1:
redirect(h, req, 302, "http://example.com/%d" % count)
redirect(h, req, "http://example.com/")
count = count + 1
except urllib2.HTTPError:
# don't stop until max_repeats, because cookies may introduce state
self.assertEqual(count, urllib2.HTTPRedirectHandler.max_repeats)
# detect endless non-repeating chain of redirects
req = Request(from_url, origin_req_host="example.com")
count = 0
try:
while 1:
redirect(h, req, "http://example.com/%d" % count)
count = count + 1
except urllib2.HTTPError:
self.assertEqual(count,
urllib2.HTTPRedirectHandler.max_redirections)
def test_cookie_redirect(self):
class MockHTTPHandler(urllib2.HTTPHandler):
def __init__(self): self._count = 0
def http_open(self, req):
import mimetools
from StringIO import StringIO
if self._count == 0:
self._count = self._count + 1
msg = mimetools.Message(
StringIO("Location: http://www.cracker.com/\r\n\r\n"))
return self.parent.error(
"http", req, MockFile(), 302, "Found", msg)
else:
self.req = req
msg = mimetools.Message(StringIO("\r\n\r\n"))
return MockResponse(200, "OK", msg, "", req.get_full_url())
# cookies shouldn't leak into redirected requests
from cookielib import CookieJar
from urllib2 import build_opener, HTTPHandler, HTTPError, \
HTTPCookieProcessor
from test_cookielib import interact_netscape
cj = CookieJar()
interact_netscape(cj, "http://www.example.com/", "spam=eggs")
hh = MockHTTPHandler()
cp = HTTPCookieProcessor(cj)
o = build_opener(hh, cp)
o.open("http://www.example.com/")
self.assert_(not hh.req.has_header("Cookie"))
class MiscTests(unittest.TestCase):

75
Lib/urllib2.py
View File

@ -106,6 +106,7 @@ import sys
import time
import urlparse
import bisect
import cookielib
try:
from cStringIO import StringIO
@ -176,7 +177,8 @@ class GopherError(URLError):
class Request:
def __init__(self, url, data=None, headers={}):
def __init__(self, url, data=None, headers={},
origin_req_host=None, unverifiable=False):
# unwrap('<URL:type://host/path>') --> 'type://host/path'
self.__original = unwrap(url)
self.type = None
@ -188,6 +190,10 @@ class Request:
for key, value in headers.items():
self.add_header(key, value)
self.unredirected_hdrs = {}
if origin_req_host is None:
origin_req_host = cookielib.request_host(self)
self.origin_req_host = origin_req_host
self.unverifiable = unverifiable
def __getattr__(self, attr):
# XXX this is a fallback mechanism to guard against these
@ -242,6 +248,12 @@ class Request:
self.host, self.type = host, type
self.__r_host = self.__original
def get_origin_req_host(self):
return self.origin_req_host
def is_unverifiable(self):
return self.unverifiable
def add_header(self, key, val):
# useful for something like authentication
self.headers[key.capitalize()] = val
@ -254,6 +266,15 @@ class Request:
return bool(header_name in self.headers or
header_name in self.unredirected_hdrs)
def get_header(self, header_name, default=None):
return self.headers.get(
header_name,
self.unredirected_hdrs.get(header_name, default))
def header_items(self):
hdrs = self.unredirected_hdrs.copy()
hdrs.update(self.headers)
return hdrs.items()
class OpenerDirector:
def __init__(self):
@ -460,7 +481,11 @@ class HTTPDefaultErrorHandler(BaseHandler):
raise HTTPError(req.get_full_url(), code, msg, hdrs, fp)
class HTTPRedirectHandler(BaseHandler):
# maximum number of redirections before assuming we're in a loop
# maximum number of redirections to any single URL
# this is needed because of the state that cookies introduce
max_repeats = 4
# maximum total number of redirections (regardless of URL) before
# assuming we're in a loop
max_redirections = 10
def redirect_request(self, req, fp, code, msg, headers, newurl):
@ -481,7 +506,10 @@ class HTTPRedirectHandler(BaseHandler):
# from the user (of urllib2, in this case). In practice,
# essentially all clients do redirect in this case, so we
# do the same.
return Request(newurl, headers=req.headers)
return Request(newurl,
headers=req.headers,
origin_req_host=req.get_origin_req_host(),
unverifiable=True)
else:
raise HTTPError(req.get_full_url(), code, msg, headers, fp)
@ -490,10 +518,12 @@ class HTTPRedirectHandler(BaseHandler):
# have already seen. Do this by adding a handler-specific
# attribute to the Request object.
def http_error_302(self, req, fp, code, msg, headers):
# Some servers (incorrectly) return multiple Location headers
# (so probably same goes for URI). Use first header.
if 'location' in headers:
newurl = headers['location']
newurl = headers.getheaders('location')[0]
elif 'uri' in headers:
newurl = headers['uri']
newurl = headers.getheaders('uri')[0]
else:
return
newurl = urlparse.urljoin(req.get_full_url(), newurl)
@ -506,20 +536,16 @@ class HTTPRedirectHandler(BaseHandler):
return
# loop detection
# .redirect_dict has a key (url, code) if url was previously
# visited as a result of a redirection with that code. The
# code is needed in addition to the URL because visiting a URL
# twice isn't necessarily a loop: there is more than one way
# to redirect (301, 302, 303, 307, refresh).
key = (newurl, code)
# .redirect_dict has a key url if url was previously visited.
if hasattr(req, 'redirect_dict'):
visited = new.redirect_dict = req.redirect_dict
if key in visited or len(visited) >= self.max_redirections:
if (visited.get(newurl, 0) >= self.max_repeats or
len(visited) >= self.max_redirections):
raise HTTPError(req.get_full_url(), code,
self.inf_msg + msg, headers, fp)
else:
visited = new.redirect_dict = req.redirect_dict = {}
visited[key] = None
visited[newurl] = visited.get(newurl, 0) + 1
# Don't close the fp until we are sure that we won't use it
# with HTTPError.
@ -912,7 +938,7 @@ class AbstractHTTPHandler(BaseHandler):
def set_http_debuglevel(self, level):
self._debuglevel = level
def do_request(self, request):
def do_request_(self, request):
host = request.get_host()
if not host:
raise URLError('no host given')
@ -987,7 +1013,7 @@ class HTTPHandler(AbstractHTTPHandler):
def http_open(self, req):
return self.do_open(httplib.HTTPConnection, req)
http_request = AbstractHTTPHandler.do_request
http_request = AbstractHTTPHandler.do_request_
if hasattr(httplib, 'HTTPS'):
class HTTPSHandler(AbstractHTTPHandler):
@ -995,7 +1021,24 @@ if hasattr(httplib, 'HTTPS'):
def https_open(self, req):
return self.do_open(httplib.HTTPSConnection, req)
https_request = AbstractHTTPHandler.do_request
https_request = AbstractHTTPHandler.do_request_
class HTTPCookieProcessor(BaseHandler):
def __init__(self, cookiejar=None):
if cookiejar is None:
cookiejar = CookieJar()
self.cookiejar = cookiejar
def http_request(self, request):
self.cookiejar.add_cookie_header(request)
return request
def http_response(self, request, response):
self.cookiejar.extract_cookies(response, request)
return response
https_request = http_request
https_response = http_response
class UnknownHandler(BaseHandler):
def unknown_open(self, req):

4
Misc/NEWS
View File

@ -311,6 +311,10 @@ Extension modules
Library
-------
- Added a new module: cookielib. Automatic cookie handling for HTTP
clients. Also, support for cookielib has been added to urllib2, so
urllib2.urlopen() can transparently handle cookies.
- stringprep.py now uses built-in set() instead of sets.Set().
- Bug #876278: Unbounded recursion in modulefinder