The following table describes the keywords to be used in a query to describe the types of outer joins available with CA Dataquery. These keywords should be used in place of the RELATED keyword.
Rows that do not form an inner equijoin are included in the active found set along with the rows joined with an inner equijoin (as in RELATED BY). They are extended to the length of other rows by adding null values. If an unmatched row is in the first table named, the null values are added to the right side and the row is considered left-joined.
If an unmatched row is in the second table named, the null values are added to the left side and the row is considered to be right-joined.
The output rows consist of all joined rows and the rows not retrieved from TABLE1, extended with null values. No unmatched rows from TABLE2 are included.
The output rows consist of all inner equijoin rows and rows not retrieved from TABLE2, extended with null values. No unmatched rows from TABLE1 are included.
The output rows consist of only rows that would be left after an inner equijoin, extended with null values. The rows from TABLE1 are extended with nulls on the right and the rows from TABLE2 are extended with nulls on the left.
The active found set consists of only unmatched rows from TABLE1.
The active found set consists of only unmatched rows from TABLE2.
Note: The use of OUTER JOIN, RIGHT-JOIN, or RIGHT-DISJOIN adds additional processing time to FIND statement processing. Outer joins and disjunctions cause the CA Dataquery optimization process to be bypassed. Any query containing an outer join or disjoin keyword is processed by accessing the tables in the order they appear in the query. Each table named on the "right side" must be read twice to accomplish the join. The use of LEFT-JOIN or LEFT-DISJOIN does not cause this additional processing.
|
Copyright © 2014 CA.
All rights reserved.
|
|