Remy Maucherat wrote:
[EMAIL PROTECTED] wrote:
jfarcand 2003/12/02 08:27:11
Modified: coyote/src/java/org/apache/coyote Request.java ActionCode.java catalina/src/share/org/apache/coyote/tomcat5 CoyoteRequest.java http11/src/java/org/apache/coyote/http11 Http11Processor.java Log: Implement getLocalPort using ActionCode
instead of getServerPort. Associate 1 ActionCode for each getXXX
method.
Please review.
I'll have to vote -0 for implementing getLocalPort as mandated in the
specification, since I think this is a huge mistake, and the spec
authors intention is misinterpreted (or they didn't think about all the consequences of the wording they used, thinking only about the case of a server without any proxies, or using an AJP-like scheme). BTW, I also think your patch has a sky high likelihood of breaking the JK 2 connector ;-)
Can you explain how? The only changes I did was to remove the code associated with remoteAddr (which was duplicated and useless IMO).
I don't see how JK would set the localPort field (since the new action isn't handled yet), so I thought it would always return -1. Maybe I'm wrong, I didn't try it.
Make sense.That will also apply to LocalAddr too then. What the specs states is:
public java.lang.String getLocalAddr() Returns the Internet Protocol (IP) address of the interface on which the request was received.
public int getLocalPort() Returns the Internet Protocol (IP) port number of the interface on which the request was received.
I think parsing/using the Host header is the way to go, since it will work with/without proxy. Can I -1 on myself ;-)
I don't know, maybe you can talk to the Servlet spec people first to see what they have to say on this issue ?
Rémy
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
