Remapping source or target columns

You can change columns and the hosts, databases, schemas, or tables for mapped columns. You might remap to change from candidate columns to physical columns or from one data source to another. When you perform a remap, the transformation operators and business terms that are part of the mapping specification remain with the remapped columns.

1. Within the Mapping Editor, select columns that you want to remap. You can select columns in the following ways:
   • Highlight columns. Click and then drag your cursor over multiple columns rows or press the Shift key or Ctrl key and click each column.
   • Right-click and select one of the following actions:

   Option	Description
   Select all from the same host	Highlights all columns that share the same host. The columns highlighted can be contained within different databases, schemas, and tables.
   Select all from the same database	Highlights all columns that share the same database. The columns highlighted can be contained within different schemas and tables.
   Select all from the same schema	Highlights all columns that share the same schema. The columns highlighted can be contained within different tables.
   Select all from the same table	Highlights all columns that share the same table.

2. With the columns highlighted, right-click and select Remap...
3. In the Remap Data Sources table, within the Matching Replacement column, change the names for columns to the replacement column names. You can drag replacement columns from the Browser tab or use the Discover tab below the Remap Data Sources table and find good replacement matches.

   Option	Description
   To remap at the host, database, schema or table level	Click the Browser tab. In the Column list, you see a list of data sources. Select a database, schema, or table and drag it into the cell under Matching Replacement that you want to replace. The multipart names for the columns change, reflecting the replacement data source.
   To remap individual columns	Click the Browser tab. In the Column list, you see a list of data sources. Select a column and drag it into the cell under Matching Replacement that you want to replace. The multipart name for the column changes.
   To search for replacements and select good matches	In the Matching Replacement column, select the columns that you want to find replacement matches for. In the Discover tab below the Remap Data Sources table, expand the folders and select the data source that you want to search within. Optional: Click Options... to view characteristics or to specify characteristics that you want to match. Click Discover Replacement. The message Results available is displayed in the cells of the Matching Replacement column. You can either update all objects with the best matches or accept replacements one at a time To update all objects for which results are available, in the Matching Replacement column, right-click and select Accept Best Match. To review match results one at a time and select a replacement, perform the following steps: In the Matching Replacement column, click a cell that contains the Results available message to see the results in the Discover tab. The resulting list of columns are sorted by their score. Tip: To see details about the match characteristics for a column, select the candidate from the list of recommended columns and click Match Details. In the Discover tab, select the column that you want and click Update. The Matching Replacement cell is updated and the next available result is highlighted.

4. Click Finish when you have completed selecting replacements. You can see that the mapping specification contains the columns that replaced in the source or target columns that were in place before the remapping.

Example

In the following example, a candidate table is replaced with a table that physically exists in the metadata repository. Then, the discover function is used to find replacements for columns that were not matched by the initial table replacement.

In the mapping editor, the user highlights candidate columns to be replaced and then right-clicks to select Remap.

Figure 1. In the mapping editor, select the columns that are to be replaced. Right-click to select Remap.

The view changes. In the Remap Data Sources table, the user can see the current data source. The user then selects the BANK2.CUSTOMERS table from the Browser tab view and drags the table to the Matching Replacement column. The user drops the table in the row of the table to be replaced.

Figure 2. From the Database metadata tab, select and drag the replacement object to the Remap Data Sources table.

When the table is dropped into the cell, the columns reflect the change. Not all columns are matched and replaced with columns from the BANK2.CUSTOMERS table. Some of the candidate columns do not have an equivalent in the BANK2.CUSTOMERS metadata information.

Figure 3. The candidate table, MASTER matched with the replacement table BANK2.CUSTOMERS.

To discover replacements for columns that remain unmatched, the user highlights the unmatched rows. When the rows are highlighted, the user can see the Discover tab below the table. The user selects the data sources that are to be searched for replacement matches.

Figure 4. Highlight unmatched columns and then use Discover to search for replacement matches.

The user clicks Discover Replacement to begin the search for replacement matches. When the discover completes, in the Remap Data Sources table, the user can see that matches were found.

Figure 5. Replacement matches are discovered and the results are available.

In the Discover Tab, the user can see the matches that were found. The user chooses to accept all of the best matches that were discovered. With the result rows highlighted, the user right-clicks and selects Accept Best Match.

Figure 6. Accept the best matches that were discovered.

The replacement is complete. Discover found columns in the BANK.CUSTOMERS table that best matched the columns that had remained unmatched after the initial remap from MASTER to BANK2.CUSTOMERS.

Figure 7. Discover found columns in the BANK.CUSTOMERS table that best matched the remaining unmatched columns.