TABLE FILE T1
PRINT *
ON TABLE HOLD AS H0 FORMAT FOCUS INDEX SITEID
END

DEFINE FILE T2
ID1/A1 = ID;
END
TABLE FILE T2
PRINT *
BY ID1
ON TABLE HOLD AS H1 FORMAT FOCUS INDEX SITEID
END

DEFINE FILE T3
ID1/A1 = ID;
END
TABLE FILE T3
PRINT *
BY ID1
ON TABLE HOLD AS H2 FORMAT FOCUS INDEX SITEID
END


JOIN
SID IN H0 TO ALL
SID IN H1 AS J1
END
JOIN
SID AND ID1 IN H0 TO ALL
SID AND ID1 IN H2 AS J2
END
TABLE FILE H0
PRINT
FIELD1
FIELD2
FIELD3
END

This code is giving me error saying LINKED FILE DOES NOT HAVE A MATCHING KEY FIELD OR SEGMENT: H2

Can some one help me?
I have tried in various ways but same error.

Thanks a lot..

This message has been edited. Last edited by: Kerry,

If your HOLD files are indexed by SITEID that's the field you should use to join, instead of SID.

Please see [SOLVED] Join Error Message

"two individual fields can be joined from the host file, the cross-referenced file requires the one concatenated field".

Why are you adding an index on SITEID and then not using it in you joins? You should be joining to an indexed field in a FOCUS db. It also might help to specify the tables you are using in you join so as not to confuse WF. You probably want

JOIN H1.SID AND H1.ID1 IN H0 TO ALL SID AND ID1 IN H2 AS J2

A couple of suggestions:

1. It is usually "good" to stay away from using PRINT *. It is preferred to list the specific fields you want to use in your request. If you only need FIELD1, FIELD2 and FIELD3 then it is better to make your request specific to that. PRINT * usually leaves some open Enigma as to what the request is all about

2. You should only need an INDEX for those files that will be used as "cross-referenced" ones. Therefore, if H0 is your host file as I can see from the sample code you posted, then an index on SITEID is not required.