The STANDARIZATION OUTPUT operator performs an action that is a special case of standardization.
The regular STANDARIZATION allows a value to be transformed based on a universal identifier contained in one of the columns of the input data stream. The new value is looked up using the universal identifier in an independently retrieved dictionary.
The STANDARIZATION OUTPUT operator allows a data value (column) to be transformed based on a dictionary that is automatically created ad hoc by the 
OUTPUT SQL operator while the data stream is being written to the database. The OUTPUT SQL operator can be configured so that certain data (columns) change their values—for example, this is the case with unique identifiers of universal objects when data comes from multiple sources. These changes are stored in a recoding table created by GRAVITY in the target database and are accessible via the 
STANDARIZATION OUTPUT operator. In other words, this can be thought of as standardization performed not against an external 
END OF PATH dictionary, but against a recoding dictionary generated by the OUTPUT SQL operator.
An example of a GRAVITY project using the STANDARIZATION OUTPUT operator:
OPERATOR CONFIGURATION
As a first step, you must select the OUTPUT SQL (there may be multiple such operators in the GRAVITY application) that is responsible for creating the recoding dictionary and the target table to be standardized during this action. GRAVITY will suggest only those tables that participate in the selected operator’s action. You must also define how GRAVITY should behave if an error is detected during standardization (for example, if an entry is missing from the recoding dictionary).
You can choose the “stop execution” action in GRAVITY — in this case, select Stop with error — or choose “continue”, in which case NULL should be assigned as the value of the processed data.
In the OUTPUT SQL PK tab, you must map a column from the input data bus to the corresponding column in the dictionary created by the OUTPUT SQL operator.
