scriptcomp: add initializers to for, while, switch statements

[tinygiant98]

This change adds the ability to use initialization and arbitrary statements within for, while and switch statements. Any variables defined or initialized within a for, while or switch statement will be scoped within the associated compound statement and any child scopes.

All legacy constructs are supported. Initialization statements are optional in for, while or switch statements.

Testing

Simple initializer in for statement:

    for (int i = 0; i < 5; i++)
    {
        Debug(c("    > iteration " + _i(i), COLOR_GREEN_LIGHT));
    }

Initializer/declaration list and comma-separated increment statements in for statement:

    for (int x, y = 10, z = -1; y > z; y--, z++, x++)
    {
        Debug(c("    > iteration " + _i(x), COLOR_GREEN_LIGHT));
        Debug(c("      > x = " + _i(x) + ", y = " + _i(y) + ", z = " + _i(z), COLOR_RED_LIGHT));
    }

Function calls in initializer and increments statements in for statement:

    for (object o = GetFirstPC(); GetIsObjectValid(o); o = GetNextPC())
    {
        Debug(c("    > pc found -> " + GetName(o), COLOR_GREEN_LIGHT));
    }

Function calls in nested for loops:

    for (object oArea = GetFirstArea(); GetIsObjectValid(oArea); oArea = GetNextArea())
    {
        Debug(c("    > area found -> " + GetName(oArea), COLOR_GREEN_LIGHT));
        for (object o = GetFirstObjectInArea(oArea); GetIsObjectValid(o); o = GetNextObjectInArea(oArea))
        {
            Debug(c("      > object found -> " + GetName(o) + " [" + GetTag(o) + "]", COLOR_ORANGE));
            if (GetObjectType(o) == OBJECT_TYPE_CREATURE)
            {
                for (effect e = GetFirstEffect(o); GetIsEffectValid(e); e = GetNextEffect(o))
                {
                    Debug(c("        > effect found -> " + _i(GetEffectType(e)), COLOR_RED_LIGHT));
                }
            }
        }
    }

Arbitrary statements in initializers and incrementers in for statement:

    int x, y;
    object o = GetFirstPC();
    for (SendMessageToPC(o, "    Loop Init"), y = 10; x < y; SendMessageToPC(o, "        Loop Iteration"), x++, y--)
    {
        SendMessageToPC(o, "      Loop Body");
        SendMessageToPC(o, "        x = " + _i(x) + ", y = " + _i(y));
    }

Simple initializer in while statement:

    while (int x = 7; x > 0)
    {
        Debug(c("    > x = " + _i(x--), COLOR_GREEN_LIGHT));
    }

Function call in initializer for switch statement

    // SWITCH_INIT = 5
    switch (int x = GetLocalInt(GetFirstPC(), "SWITCH_INIT"); x)
    {
        case 0: Debug(c("    > 0")); break;
        case 1: Debug(c("    > 1")); break;
        case 2: Debug(c("    > 2")); break;
        case 3: Debug(c("    > 3")); break;
        case 4: Debug(c("    > 4")); break;
        case 5: Debug(c("    > 5")); break;
    }

Changelog

Added

• Added initializer statements to for, while and switch statements.
• Added comma-separated increment statements to for statements.

Licence

• I am licencing my change under the project's MIT licence, including all changes to GPL-3.0 licenced parts of the codebase.

[Daztek]

How does it handle nested loops that use the same variable name? Are they scoped properly?

for (int x = 0; x < 100; x++)
{
    for (int x = 0; x < 10; x++)
    {
        
    }
}

[mtijanic]

Just leaving a few nits here while I actually make sense of the bigger change. This will take a while.

[tinygiant98]

How does it handle nested loops that use the same variable name? Are they scoped properly?

for (int x = 0; x < 100; x++)
{
    for (int x = 0; x < 10; x++)
    {
        
    }
}

Ran a nested scope test. Results are on discord, but I'll put there here for record-keeping purposes.

Test code snippet:

for (int x = 0; x < 5; x++)
{
    Debug(c("  start outer loop -> x = " + _i(x), COLOR_ORANGE));
    for (int x = 3; x > 0; x--)
    {
        Debug(c("    inner loop -> x = " + _i(x), COLOR_RED_LIGHT));
    }
    Debug(c("  end outer loop -> x = " + _i(x), COLOR_GREEN_LIGHT));
}

I think it meets your intent, but if this test is too simple, let me know.

[tinygiant98]

I've run all the tests I can think of here, but I also know how it's supposed to work, so I'm probably missing some ideas on how to break it. If you guys have any tests you'd like run against this PR, let me know and I'll run 'em.