Basic question about using || in script calls

[cuby]

Hi all, very quick question.

I have a conditional branch that is checking on the value of a variable. The variable is is a text string, and I am checking this string to see if it says a certain thing before proceeding. However, I'm having some issues with this and since I'm a newbie in this area, I trust I'm doing something wrong and hoping to get some help.

Right now in my conditional branch I'm checking script:

$gameVariables.value(1) == "blah" || "blah2" || "blah3"

If var 1 matches either of those three, hen go to Label "X", else exit the event. However, no matter what the variable is, it's going to Label X. I've done a quick test to confirm that no matter the value of var 1, this test is passing as true.

Could anyone say what I'm doing wrong? Thank you!

c

 

[Bex]

Should it not be written === instead of ==

 

[Hisao Shou]

Hi all, very quick question.

I have a conditional branch that is checking on the value of a variable. The variable is is a text string, and I am checking this string to see if it says a certain thing before proceeding. However, I'm having some issues with this and since I'm a newbie in this area, I trust I'm doing something wrong and hoping to get some help.

Right now in my conditional branch I'm checking script:

$gameVariables.value(1) == "blah" || "blah2" || "blah3"

If var 1 matches either of those three, hen go to Label "X", else exit the event. However, no matter what the variable is, it's going to Label X. I've done a quick test to confirm that no matter the value of var 1, this test is passing as true.

Could anyone say what I'm doing wrong? Thank you!

c

You should better make the check separately like this:

$gameVariables.value(1) == "blah" || $gameVariables.value(1) == "blah2" || $gameVariables.value(1) == "blah 3"

 

[cuby]

You should better make the check separately like this:

$gameVariables.value(1) == "blah" || $gameVariables.value(1) == "blah2" || $gameVariables.value(1) == "blah 3"

Thanks, this worked!!

 

[caethyril]

To explain a bit:

1. The === operator only ever checks one pair of values, i.e. A === B.
2. Equality comparison also has higher precedence than logical OR, i.e. === evaluates before ||.
3. JavaScript treats any non-empty string as "truthy".

Putting it all together, this condition:

---

$gameVariables.value(1) == "blah" || "blah2" || "blah3"

---

...will reduce to this every time:

---

something OR true OR true

---

...which is always true.

In case it helps, here's a slightly more concise way of writing your 3-value comparison using Array#indexOf:

["blah", "blah2", "blah3"].indexOf($gameVariables.value(1)) >= 0;

I.e. "is the value of game variable 1 found in the list ["blah", "blah2", "blah3"]?"

Is this thread solved?

 

[Trihan]

One thing that might benefit from elaboration is == vs ===, since both work but slightly differently.

== is an equality operator that does type coercion; that is to say, if there's a way to make the two things being compared the same type, the interpreter will do so for you. So 1 == "1" will evaluate to true, since they're both converted to the same type before comparison. === is a type-strict equality operator that does not perform coercion, so 1 === "1" returns false because one is a Number and one is a String.

 

[caethyril]

Whoops, yes, I use === so much that I didn't notice the original post used ==. The logic I described still applies, except that indexOf uses strict equality, ===.

 

[Trihan]

Whoops, yes, I use === so much that I didn't notice the original post used ==. The logic I described still applies, except that indexOf uses strict equality, ===.

I could just console.time it, but since you're here you might already know: do you know if using the indexOf method is faster than using .contains or .includes?

 

[caethyril]

@Trihan: I haven't tested, no. contains in MZ is just an alias for ES6 includes; in MV it invokes indexOf. ES6 includes might have some compiler-level optimisation, but there's probably not much difference, especially for small arrays.

 

[Trihan]

I know contains is just an alias. The backfill for includes is apparently

Array.includes = function(val,arr) {
    for(i=0;i<arr.length;i++){
       if (!ret || ret == false){ret = val == arr[i] ? true: false;}
    }
    return ret;
  }

which doesn't use indexOf. Obviously that's probably not the same native code as ES6's implementation, though. I might console.time it and see which is faster.

 

[caethyril]

@Trihan: I don't think I'd use that for a polyfill! ret == false will never be checked; == returns a boolean so the ? true : false is redundant; no early break when val is found; no support for the fromIndex parameter; ret should be declared locally.

Also, a nit-pick: ES6 is a specification, the implementation is browser-dependent.

I just ran a few tests using some of my in-progress code (see spoiler) in a new RMMZ project via the console. Repeat 10000-trial tests using indexOf vs includes to search the same, ordered 1000-element array for a random value yielded results of both around 13~16 ms per approach (excluding setup), neither consistently faster than the other. Your results may vary, but I'd say it's OK to use whichever you prefer.

Spoiler: test code

Test method:

const test = function(fArr, setup, trials = 10000, labels = [], sFigs = 6) {
    // Check
    if (!Array.isArray(fArr) || fArr.some(
            f => typeof f !== 'function'
        )) return null;
    // Initialise
    const L = fArr.length;
    let t  = 0;
    let tS = 0;
    let tF = (new Array(L)).fill(0);
    // Measure
    let tT = -performance.now();
    for (let n = Math.max(trials, 0); n--;) {
        t = performance.now();
        const data = setup?.();
        tS += performance.now() - t;
        for (let m = L; m--;) {
            t = performance.now();
            fArr[m](data);
            tF[m] += performance.now() - t;
        }
    }
    tT += performance.now();
    // Format
    const [
        lF = 'fn',
        lS = 'setup',
        lO = 'other',
        lT = 'total',
        pt = 'time (ms)',
        pp = 'time %'
    ] = labels;
    const res = {};
    for (let n = 0; n < L; n++) {
        const k = `${lF}${n} ${fArr[n].name}`.trim();
        res[k] = tF[n];
    }
    res[lS] = tS;
    res[lO] = tT - tS - tF.reduce((a, c) => a + c, 0);
    const mag = v => 10 ** (sFigs - 1 - Math.floor(Math.log10(v)));
    for (const [k, v] of Object.entries(res)) {
        const ord = mag(v);
        res[k] = {
            [pt]: Math.round(ord * v) / ord,
            [pp]: Math.round(ord * 100 * v / tT) / ord
        };
    }
    (ord => res[lT] = { [pt]: Math.round(ord * tT) / ord, [pp]: 100 })(mag(tT));
    // Present
    SceneManager.showDevTools();
    console.table(res);
    return res;
};

Test run:

test(
    [       // tests
        arr => arr.indexOf(Math.randomInt(1000)) >= 0,
        arr => arr.includes(Math.randomInt(1000))
    ],
    () => { // setup
        const a = new Array(1000);
        for (var n = 1000; n--;)
            a[n] = n;
        return a;
    }
);

(Sorry for derailing~ )

 

[Trihan]

@Trihan: I don't think I'd use that for a polyfill! ret == false will never be checked; == returns a boolean so the ? true : false is redundant; no early break when val is found; no support for the fromIndex parameter; ret should be declared locally.

Also, a nit-pick: ES6 is a specification, the implementation is browser-dependent.

I just ran a few tests using some of my in-progress code (see spoiler) in a new RMMZ project via the console. Repeat 10000-trial tests using indexOf vs includes to search the same, ordered 1000-element array for a random value yielded results of both around 13~16 ms per approach (excluding setup), neither consistently faster than the other. Your results may vary, but I'd say it's OK to use whichever you prefer.

Spoiler: test code

Test method:

const test = function(fArr, setup, trials = 10000, labels = [], sFigs = 6) {
    // Check
    if (!Array.isArray(fArr) || fArr.some(
            f => typeof f !== 'function'
        )) return null;
    // Initialise
    const L = fArr.length;
    let t  = 0;
    let tS = 0;
    let tF = (new Array(L)).fill(0);
    // Measure
    let tT = -performance.now();
    for (let n = Math.max(trials, 0); n--;) {
        t = performance.now();
        const data = setup?.();
        tS += performance.now() - t;
        for (let m = L; m--;) {
            t = performance.now();
            fArr[m](data);
            tF[m] += performance.now() - t;
        }
    }
    tT += performance.now();
    // Format
    const [
        lF = 'fn',
        lS = 'setup',
        lO = 'other',
        lT = 'total',
        pt = 'time (ms)',
        pp = 'time %'
    ] = labels;
    const res = {};
    for (let n = 0; n < L; n++) {
        const k = `${lF}${n} ${fArr[n].name}`.trim();
        res[k] = tF[n];
    }
    res[lS] = tS;
    res[lO] = tT - tS - tF.reduce((a, c) => a + c, 0);
    const mag = v => 10 ** (sFigs - 1 - Math.floor(Math.log10(v)));
    for (const [k, v] of Object.entries(res)) {
        const ord = mag(v);
        res[k] = {
            [pt]: Math.round(ord * v) / ord,
            [pp]: Math.round(ord * 100 * v / tT) / ord
        };
    }
    (ord => res[lT] = { [pt]: Math.round(ord * tT) / ord, [pp]: 100 })(mag(tT));
    // Present
    SceneManager.showDevTools();
    console.table(res);
    return res;
};

Test run:

test(
    [       // tests
        arr => arr.indexOf(Math.randomInt(1000)) >= 0,
        arr => arr.includes(Math.randomInt(1000))
    ],
    () => { // setup
        const a = new Array(1000);
        for (var n = 1000; n--;)
            a[n] = n;
        return a;
    }
);

(Sorry for derailing~ )

That's fair, I just found the apparent polyfill in a SO post, I never actually read the code.

 

[cuby]

In case it helps, here's a slightly more concise way of writing your 3-value comparison using Array#indexOf:

["blah", "blah2", "blah3"].indexOf($gameVariables.value(1)) >= 0;

I.e. "is the value of game variable 1 found in the list ["blah", "blah2", "blah3"]?"

Is this thread solved?

I totally get it now, thank you very much for explaining it so clearly.

The array you suggested is very obviously the optimal solution, since I actually have lots and lots of values to check against, and this is surely the most efficient. Let me just run a quick test of it, though as with everything you suggest, I'm sure it works

EDIT: Confirmed, works. You're the best @caethyril. Solved!

 

[caethyril]

Great! Happy RPG Making~

This thread is being closed, due to being solved. If for some reason you would like this thread re-opened, please report this post and leave a message why. Thank you.