cookiespecbase.java

爬虫

      * @param port the port from which the {@link Cookie} was received
      * @param path the path from which the {@link Cookie} was received
      * @param secure <tt>true</tt> when the {@link Cookie} was received using a
      * secure connection
      * @param cookie The cookie to validate.
      * @throws MalformedCookieException if an exception occurs during
      * validation
      */
    
    public void validate(String host, int port, String path, 
        boolean secure, final Cookie cookie) 
        throws MalformedCookieException {
            
        LOG.trace("enter CookieSpecBase.validate("
            + "String, port, path, boolean, Cookie)");
        if (host == null) {
            throw new IllegalArgumentException(
                "Host of origin may not be null");
        }
        if (host.trim().equals("")) {
            throw new IllegalArgumentException(
                "Host of origin may not be blank");
        }
        if (port < 0) {
            throw new IllegalArgumentException("Invalid port: " + port);
        }
        if (path == null) {
            throw new IllegalArgumentException(
                "Path of origin may not be null.");
        }
        if (path.trim().equals("")) {
            path = PATH_DELIM;
        }
        host = host.toLowerCase();
        // check version
        if (cookie.getVersion() < 0) {
            throw new MalformedCookieException ("Illegal version number " 
                + cookie.getValue());
        }

        // security check... we musn't allow the server to give us an
        // invalid domain scope

        // Validate the cookies domain attribute.  NOTE:  Domains without 
        // any dots are allowed to support hosts on private LANs that don't 
        // have DNS names.  Since they have no dots, to domain-match the 
        // request-host and domain must be identical for the cookie to sent 
        // back to the origin-server.
        if (host.indexOf(".") >= 0) {
            // Not required to have at least two dots.  RFC 2965.
            // A Set-Cookie2 with Domain=ajax.com will be accepted.

            // domain must match host
            if (!host.endsWith(cookie.getDomain())) {
                String s = cookie.getDomain();
                if (s.startsWith(".")) {
                    s = s.substring(1, s.length());
                }
                if (!host.equals(s)) { 
                    throw new MalformedCookieException(
                        "Illegal domain attribute \"" + cookie.getDomain() 
                        + "\". Domain of origin: \"" + host + "\"");
                }
            }
        } else {
            if (!host.equals(cookie.getDomain())) {
                throw new MalformedCookieException(
                    "Illegal domain attribute \"" + cookie.getDomain() 
                    + "\". Domain of origin: \"" + host + "\"");
            }
        }

        // another security check... we musn't allow the server to give us a
        // cookie that doesn't match this path

        if (!path.startsWith(cookie.getPath())) {
            throw new MalformedCookieException(
                "Illegal path attribute \"" + cookie.getPath() 
                + "\". Path of origin: \"" + path + "\"");
        }
    }


    /**
     * Return <tt>true</tt> if the cookie should be submitted with a request
     * with given attributes, <tt>false</tt> otherwise.
     * @param host the host to which the request is being submitted
     * @param port the port to which the request is being submitted (ignored)
     * @param path the path to which the request is being submitted
     * @param secure <tt>true</tt> if the request is using a secure connection
     * @param cookie {@link Cookie} to be matched
     * @return true if the cookie matches the criterium
     */

    public boolean match(String host, int port, String path, 
        boolean secure, final Cookie cookie) {
            
        LOG.trace("enter CookieSpecBase.match("
            + "String, int, String, boolean, Cookie");
            
        if (host == null) {
            throw new IllegalArgumentException(
                "Host of origin may not be null");
        }
        if (host.trim().equals("")) {
            throw new IllegalArgumentException(
                "Host of origin may not be blank");
        }
        if (port < 0) {
            throw new IllegalArgumentException("Invalid port: " + port);
        }
        if (path == null) {
            throw new IllegalArgumentException(
                "Path of origin may not be null.");
        }
        if (cookie == null) {
            throw new IllegalArgumentException("Cookie may not be null");
        }
        if (path.trim().equals("")) {
            path = PATH_DELIM;
        }
        host = host.toLowerCase();
        if (cookie.getDomain() == null) {
            LOG.warn("Invalid cookie state: domain not specified");
            return false;
        }
        if (cookie.getPath() == null) {
            LOG.warn("Invalid cookie state: path not specified");
            return false;
        }
        
        return
            // only add the cookie if it hasn't yet expired 
            (cookie.getExpiryDate() == null 
                || cookie.getExpiryDate().after(new Date()))
            // and the domain pattern matches 
            && (domainMatch(host, cookie.getDomain()))
            // and the path is null or matching
            && (pathMatch(path, cookie.getPath()))
            // and if the secure flag is set, only if the request is 
            // actually secure 
            && (cookie.getSecure() ? secure : true);      
    }

    /**
     * Performs domain-match as implemented in common browsers.
     * @param host The target host.
     * @param domain The cookie domain attribute.
     * @return true if the specified host matches the given domain.
     */
    public boolean domainMatch(final String host, String domain) {
        if (host.equals(domain)) {
            return true;
        }
        if (!domain.startsWith(".")) {
            domain = "." + domain;
        }
        return host.endsWith(domain) || host.equals(domain.substring(1));
    }

    /**
     * Performs path-match as implemented in common browsers.
     * @param path The target path.
     * @param topmostPath The cookie path attribute.
     * @return true if the paths match
     */
    public boolean pathMatch(final String path, final String topmostPath) {
        boolean match = path.startsWith (topmostPath);
        // if there is a match and these values are not exactly the same we have
        // to make sure we're not matcing "/foobar" and "/foo"
        if (match && path.length() != topmostPath.length()) {
            if (!topmostPath.endsWith(PATH_DELIM)) {
                match = (path.charAt(topmostPath.length()) == PATH_DELIM_CHAR);
            }
        }
        return match;
    }

    /**
     * Return an array of {@link Cookie}s that should be submitted with a
     * request with given attributes, <tt>false</tt> otherwise.
     * @param host the host to which the request is being submitted
     * @param port the port to which the request is being submitted (currently
     * ignored)
     * @param path the path to which the request is being submitted
     * @param secure <tt>true</tt> if the request is using a secure protocol
     * @param cookies an array of <tt>Cookie</tt>s to be matched
     * @return an array of <tt>Cookie</tt>s matching the criterium
     * 
// BEGIN IA CHANGES
     * @deprecated use match(String, int, String, boolean, SortedMap)
// END IA CHANGES
     */

    public Cookie[] match(String host, int port, String path, 
        boolean secure, final Cookie cookies[]) {
            
        LOG.trace("enter CookieSpecBase.match("
            + "String, int, String, boolean, Cookie[])");

        if (cookies == null) {
            return null;
        }
        List matching = new LinkedList();
        for (int i = 0; i < cookies.length; i++) {
            if (match(host, port, path, secure, cookies[i])) {
                addInPathOrder(matching, cookies[i]);
            }
        }
        return (Cookie[]) matching.toArray(new Cookie[matching.size()]);
    }

//  BEGIN IA CHANGES
    /**
     * Return an array of {@link Cookie}s that should be submitted with a
     * request with given attributes, <tt>false</tt> otherwise. 
     * 
     * If the SortedMap comes from an HttpState and is not itself
     * thread-safe, it may be necessary to synchronize on the HttpState
     * instance to protect against concurrent modification. 
     *
     * @param host the host to which the request is being submitted
     * @param port the port to which the request is being submitted (currently
     * ignored)
     * @param path the path to which the request is being submitted
     * @param secure <tt>true</tt> if the request is using a secure protocol
     * @param cookies SortedMap of <tt>Cookie</tt>s to be matched
     * @return an array of <tt>Cookie</tt>s matching the criterium
     */

    public Cookie[] match(String host, int port, String path, 
        boolean secure, final SortedMap cookies) {
            
        LOG.trace("enter CookieSpecBase.match("
           + "String, int, String, boolean, SortedMap)");

        // TODO: skip meaningless 'narrowing' when host is a numeric IP
        // (harmless in the meantime)
        
        if (cookies == null) {
            return null;
        }
        List matching = new LinkedList();
        String narrowHost = host;
        do {
            Iterator iter = cookies.subMap(narrowHost,
                    narrowHost + Cookie.DOMAIN_OVERBOUNDS).values().iterator();
            while (iter.hasNext()) {
                Cookie cookie = (Cookie) (iter.next());
                if (match(host, port, path, secure, cookie)) {
                    addInPathOrder(matching, cookie);
                }
            }
            StoredIterator.close(iter);
            int trimTo = narrowHost.indexOf('.', 1);
            narrowHost = (trimTo < 0) ? null : narrowHost.substring(trimTo+1);
        } while (narrowHost != null);

        return (Cookie[]) matching.toArray(new Cookie[matching.size()]); 
    }
//  END IA CHANGES
    
    /**
     * Adds the given cookie into the given list in descending path order. That
     * is, more specific path to least specific paths.  This may not be the
     * fastest algorythm, but it'll work OK for the small number of cookies
     * we're generally dealing with.
     *
     * @param list - the list to add the cookie to
     * @param addCookie - the Cookie to add to list
     */
    private static void addInPathOrder(List list, Cookie addCookie) {
        int i = 0;

        for (i = 0; i < list.size(); i++) {
            Cookie c = (Cookie) list.get(i);
            if (addCookie.compare(addCookie, c) > 0) {
                break;
            }
        }
        list.add(i, addCookie);
    }

    /**
     * Return a string suitable for sending in a <tt>"Cookie"</tt> header
     * @param cookie a {@link Cookie} to be formatted as string
     * @return a string suitable for sending in a <tt>"Cookie"</tt> header.
     */
    public String formatCookie(Cookie cookie) {
        LOG.trace("enter CookieSpecBase.formatCookie(Cookie)");
        if (cookie == null) {
            throw new IllegalArgumentException("Cookie may not be null");
        }
        StringBuffer buf = new StringBuffer();
        buf.append(cookie.getName());
        buf.append("=");
        String s = cookie.getValue();
        if (s != null) {
            buf.append(s);
        }
        return buf.toString();
    }

    /**
     * Create a <tt>"Cookie"</tt> header value containing all {@link Cookie}s in
     * <i>cookies</i> suitable for sending in a <tt>"Cookie"</tt> header
     * @param cookies an array of {@link Cookie}s to be formatted
     * @return a string suitable for sending in a Cookie header.
     * @throws IllegalArgumentException if an input parameter is illegal
     */

    public String formatCookies(Cookie[] cookies)
      throws IllegalArgumentException {
        LOG.trace("enter CookieSpecBase.formatCookies(Cookie[])");
        if (cookies == null) {
            throw new IllegalArgumentException("Cookie array may not be null");
        }
        if (cookies.length == 0) {
            throw new IllegalArgumentException("Cookie array may not be empty");
        }

        StringBuffer buffer = new StringBuffer();
        for (int i = 0; i < cookies.length; i++) {
            if (i > 0) {
                buffer.append("; ");
            }
            buffer.append(formatCookie(cookies[i]));
        }
        return buffer.toString();
    }


    /**
     * Create a <tt>"Cookie"</tt> {@link Header} containing all {@link Cookie}s
     * in <i>cookies</i>.
     * @param cookies an array of {@link Cookie}s to be formatted as a <tt>"
     * Cookie"</tt> header
     * @return a <tt>"Cookie"</tt> {@link Header}.
     */
    public Header formatCookieHeader(Cookie[] cookies) {
        LOG.trace("enter CookieSpecBase.formatCookieHeader(Cookie[])");
        return new Header("Cookie", formatCookies(cookies));
    }


    /**
     * Create a <tt>"Cookie"</tt> {@link Header} containing the {@link Cookie}.
     * @param cookie <tt>Cookie</tt>s to be formatted as a <tt>Cookie</tt>
     * header
     * @return a Cookie header.
     */
    public Header formatCookieHeader(Cookie cookie) {
        LOG.trace("enter CookieSpecBase.formatCookieHeader(Cookie)");
        return new Header("Cookie", formatCookie(cookie));
    }

}