Search code examples
performancestring-concatenationopenedgeprogress-4gl

Performance of OpenEdge concatenation

I have noticed in Progress (11.4) that repeated string concatenation can be very slow. For example in the following code.

DEF VAR i AS INT NO-UNDO. 
DEF VAR c AS LONGCHAR NO-UNDO.

DO  i = 1 TO 1000000:
    c = c + STRING(i MOD 10). 
END.

From my experience in Java, I think the problem is that each time we are concatenating, we are copying c which is an O(n) operation, so the whole procedure is O(n^2). Java provides the StringBuilder class to solve this problem.

Is my analysis of the problem correct? And if so, is there a solution?

Solution

  • This is a known issue - the following link also contains a work-around by buffering concatenations in character variables:

    https://blog.abevoelker.com/introducing_bigcharacter/

    Recognized by Progress and addressed in 11.7.2:

    https://knowledgebase.progress.com/articles/Article/concatenating-to-longchar-takes-exponentially-longer