Re: If boolean vs. If not boolean

On 04 Oct 2013 14:25, Glenn Gundermann wrote:

Forgive the trivialness of this question but I am wondering if you
smart folks can answer this.

Compare these two statements:

If booleanVar;
If not booleanVar;

Do both of these take the exact same number of internal instructions
for the test or does the second statement take an extra step?

It depends... On the definition of "internal instructions", among other things.

For a non-boolean variable... Typically one MI instruction effects the type-specific data comparison, and that instruction is coded with either the extension <branch-on-equal> or the extension <branch-on-unequal>.

For a boolean variable, the effect should be the same, because the RPG boolean variable is effectively just another scalar data type. The obvious benefit however [over other data types with more than a binary yes|no type of comparison\answer], is that the generated code can be a comparison against an effective *true or *false, irrespective of the "NOT" in the predicate; i.e. in the following two sets of predicates, the three predicates in each set are identical:

If BooleanVar then...
If BooleanVar='1' then...
If Not BooleanVar='0' then...

If Not BooleanVar then...
If Not BooleanVar='1' then...
If BooleanVar='0' then...

The RPG likely generates code that consistently executes the next instruction on true [or false] and then executes the branch\GOTO on false [or true]. Thus the impact is less about either what or how many instructions are generated, but about the overall percentage of actual evaluations will be to true vs to false on either form of the predicate. Even so, the evaluation which follows the non-branched path might just encounter a branch-unconditional somewhere later, to avoid whatever code exists for an else-path.