An issue with the CSV import

[miksko]

The menu option Publish->Model Exchange->CSV let you export and import items as CSV files. I've tested it with Requirements, to figure out if there are cases that don't work. There is one: If you export a requriement model, remove one of the rows and then import the CSV file again, the matching requirement isn't removed from the model.
All other kinds of updates work marvelously though. You may even add a line in the CSV file, which Sparx will interpret as a new requirement and add to the model. You will get a warning that one of the rows misses a GUID, but the application will then generate one and add the requirement to the model.

Does anyone have an idea to why removals aren't executed? This is a file repository, so it shouldn't be a permission issue.

[RobCDeJong]

The import only adds or changes elements by design.

When you want to delete elements based on an import file, you should use scripting to figure out which elements are obsolete and can be removed or modify the status, which is better for requirements management.

[miksko]

The import only adds or changes elements by design.

Thanks Rob. I had started to suspect that. The user guide mentions a lot of other prerequisites, like Type being mandatory and what will happen if you add rows without a GUID, but not this one. I came to the same conclusion as you though, that shifting State would be comparable. (Though it still leaves you with unwanted elements in diagrams, since the report filter only works on text.)

[DeBAAT]

Some tools include a column "Action" allowing you to specify what to do with that row, e.g. "Delete" could be such an action.
Another workaround might be to check the value of last modified to identify the requirements not touched by the import, thereby rendering them as optionally to be deleted.