The previous assumption that a format catalog had a primary key was.. completely wrong.
START / END / LABEL may all contain missing values, and the entire CNTLOUT dataset itself may contain duplicate rows (eg with multilabel option). The order is also important and must be preserved.
To cater for these scenarios we will need to add a new column to the exported catalog data, and use this as a key column in place of start/end/label:
FMTROW - an integer showing the position of each row, windowed by TYPE FMTNAME
Tests will be needed covering the multilabel and notsorted options.
The previous assumption that a format catalog had a primary key was.. completely wrong.
START / END / LABEL may all contain missing values, and the entire CNTLOUT dataset itself may contain duplicate rows (eg with
multilabel
option). The order is also important and must be preserved.To cater for these scenarios we will need to add a new column to the exported catalog data, and use this as a key column in place of start/end/label:
by TYPE FMTNAME
Tests will be needed covering the multilabel and notsorted options.
Affected macros: