Skip to content

Data Filtering

To select the Merged Data Filters panel, select option 6 (FILTERING) from the Profile panel. The Merged Data Filters panel (SERCLEAN), shown below, appears.

SERCLEAN ------------------- Merged Data Filters --------------------------- 
Enter > or < or <> in TYPE and           Enter a string in MATCHING TEXT and 
one or more derivative ID in MARKING     optionally a column position 

        TYPE MARKING                             MATCHING TEXT     POSITION 
         >   BC                      Or      * Notes                   7
         __  _______                 Or      _____________________     ___
         __  _______                 Or      _____________________     ___
         __  _______                 Or      _____________________     ___
         __  _______                 Or      _____________________     ___
         __  _______                 Or      _____________________     ___
         __  _______                 Or      _____________________     ___
         __  _______                 Or      _____________________     ___
         __  _______                 Or      _____________________     ___

...

The Data Filtering of the merged base and derivative versions is implemented as a logical deletion of certain records so that they will not be exported. The following are two ways to filter out unwanted records:

  • By specifying the record type and origin; for example, all records inserted by derivatives B and C are depicted as >BC; to logically delete such records, enter > in the TYPE field and BC in the MARKING field (use the Merged Data Filters panel above as an example).

  • By specifying a part of the content of the record (a string) at a certain position or anywhere within the record; for example, to delete all records that contain the string *Notes starting at position 7, enter *Notes in the MATCHING TEXT field and 7 in the POSITION field.

The data filtering is automatically done as part of the compare and merge of the versions. The logically deleted records are marked with /* on the first display of the Reconcile panel. If you like to examine the results of the data filtering, you can issue eXclude REmaining (X #) primary command. Conversely, you can issue View REmaining (V #) command to see all filtered out records disappear just as they would on export. Should you want to selectively reverse the result of data filtering, you can use the Undelete line command.

Up to 16 data filters of each category can be entered on the Merged Data Filters panel. The "Or" between a type/origin filter and a matching text filter means that if a record matches either of them it will be logically deleted. This way, you can specify up to 32 independent criteria for filtering out unwanted records from the merged versions.

The search for the text at a certain (or any) position is not case sensitive. The text can contain several words separated with blanks. If the text is enclosed within a pair of apostrophes or quotation marks, the search for that string becomes case sensitive. To search for a text anywhere on a record, leave the POSITION field blank.

The primary command RESet blanks out all merged data filters.

The Data Filtering can be used for many purposes. For example, the type/origin filters can logically delete an entire derivative, say X, by specifying the following merged data filters: <>X and > X, and setting the profile export option AutoOverride to No to avoid any deletions made in X. By using text/at position requests, many blocks of comment records can be dropped on export if unwanted.

More sophisticated applications of Merged Data Filtering may be, for example, propagations of software changes down the corporate hierarchy. Usually, the changes in the business rules come from the HQ while the branches make some local changes. In order to propagate the business rules' changes into a customized local software, the following consolidation process can be devised:

  • Define the local code as the base, the current version of the central code as a derivative, say B, and the new version of the central code as another derivative, say C.

  • All the differences between the derivatives and the base need to be filtered out except for those that originate from the new version C as they represent the changes in the business rules.

  • Thus, specify the Merged Data Filters >BC, <>BC, >B, and <>B, then set the profile export option AutoOverride to No to neutralize derivative deletions like <BC, <B, and <C.

  • After execution and export, use the auto-consolidated version as a derivative and the local code as the base to reconcile the business rules' changes with the local customizations.