Related
In JS or OOP language the polymorhpism is created by making different types.
For example:
class Field {...}
class DropdownField extends Field {
getValue() {
//implementation ....
}
}
Imagine I have library forms.js with some methods:
class Forms {
getFieldsValues() {
let values = [];
for (let f of this.fields) {
values.push(f.getValue());
}
return values;
}
}
This gets all field values. Notice the library doesnt care what field it is.
This way developer A created the library and developer B can make new fields: AutocompleterField.
He can add methods in AutocompleterField withouth changing the library code (Forms.js) .
If I use functional programming method in JS, how can I achieve this?
If I dont have methods in object i can use case statements but this violates the principle. Similar to this:
if (field.type == 'DropdownField')...
else if (field.type == 'Autocompleter')..
If developer B add new type he should change the library code.
So is there any good way to solve the issue in javascript without using object oriented programming.
I know Js isnt exactly OOP nor FP but anyway.
Thanks
JavaScript being a multi-purpose language, you can of course solve it in different ways. When switching to functional programming, the answer is really simple: Use functions! The problem with your example is this: It is so stripped down, you can do exactly the same it does with just 3 lines:
// getValue :: DOMNode -> String
const getValue = field => field.value;
// readForm :: Array DOMNode -> Array String
const readForm = formFields => formFields.map(getValue);
readForm(Array.from(document.querySelectorAll('input, textarea, select')));
// -> ['Value1', 'Value2', ... 'ValueN']
The critical thing is: How is Field::getValue() implemented, what does it return? Or more precisely: How does DropdownField::getValue() differ from AutocompleteField::getValue() and for example NumberField::getValue()? Do all of them just return the value? Do they return a pair of name and value? Do they even need to be different?
The question is therefor, do your Field classes and their inheriting classes differ because of the way their getValue() methods work or do they rather differ because of other functionality they have? For example, the "autocomplete" functionality of a textfield isn't (or shouldn't be) tied to the way the value is taken from it.
In case you really need to read the values differently, you can implement a function which takes a map/dictionary/object/POJO of {fieldtype: readerFunction} pairs:
/* Library code */
// getTextInputValue :: DOMNode -> String
const getTextInputValue = field => field.value;
// getDropdownValue :: DOMNode -> String
const getDropdownValue = field => field.options[field.selectedIndex].value;
// getTextareaValue :: DOMNode -> String
const getTextareaValue = field => field.textContent;
// readFieldsBy :: {String :: (a -> String)} -> DOMNode -> Array String
readFieldsBy = kv => form => Object.keys(kv).reduce((acc, k) => {
return acc.concat(Array.from(form.querySelectorAll(k)).map(kv[k]));
}, []);
/* Code the library consumer writes */
const readMyForm = readFieldsBy({
'input[type="text"]': getTextInputValue,
'select': getDropdownValue,
'textarea': getTextareaValue
});
readMyForm(document.querySelector('#myform'));
// -> ['Value1', 'Value2', ... 'ValueN']
Note: I intentionally didn't mention things like the IO monad here, because it would make stuff more complicated, but you might want to look it up.
In JS or OOP language the polymorhpism is created by making different types.
Yes. Or rather, by implementing the same type interface in different objects.
How can I use Javascript polymorphism without OOP classes
You seem to confuse classes with types here. You don't need JS class syntax to create objects at all.
You can just have
const autocompleteField = {
getValue() {
…
}
};
const dropdownField = {
getValue() {
…
}
};
and use the two in your Forms instance.
Depends on what you mean by "polymorphism". There's the so-called ad-hoc polymorphism which type classes in Haskell, Scala, or PureScript provide -- and this kind of dispatch is usually implemented by passing witness objects along as additional function arguments, which then will know how to perform the polymorphic functionality.
For example, the following PureScript code (from the docs), which provides a show function for some types:
class Show a where
show :: a -> String
instance showString :: Show String where
show s = s
instance showBoolean :: Show Boolean where
show true = "true"
show false = "false"
instance showArray :: (Show a) => Show (Array a) where
show xs = "[" <> joinWith ", " (map show xs) <> "]"
example = show [true, false]
It gets compiled to the following JS (which I shortened):
var Show = function (show) {
this.show = show;
};
var show = function (dict) {
return dict.show;
};
var showString = new Show(function (s) {
return s;
});
var showBoolean = new Show(function (v) {
if (v) {
return "true";
};
if (!v) {
return "false";
};
throw new Error("Failed pattern match at Main line 12, column 1 - line 12, column 37: " + [ v.constructor.name ]);
});
var showArray = function (dictShow) {
return new Show(function (xs) {
return "[" + (Data_String.joinWith(", ")(Data_Functor.map(Data_Functor.functorArray)(show(dictShow))(xs)) + "]");
});
};
var example = show(showArray(showBoolean))([ true, false ]);
There's absolutely no magic here, just some additional arguments. And at the "top", where you actually know concrete types, you have to pass in the matching concrete witness objects.
In your case, you would pass around something like a HasValue witness for different forms.
You could use a the factory pattern to ensure you follow the open close principle.
This principle says "Software entities (classes, modules, functions, etc.) should be open for extension, but closed for modification".
class FieldValueProviderFactory {
getFieldValue(field) {
return this.providers.find(p => p.type === field.type).provider(field);
}
registerProvider(type, provider) {
if(!this.providers) {
this.providers = [];
}
this.providers.push({type:type, provider:provider});
}
}
var provider = new FieldValueProviderFactory();
provider.registerProvider('DropdownField', (field) => [ 1, 2, 3 ]);
provider.registerProvider('Autocompleter', (field) => [ 3, 2, 1 ]);
class FieldCollection {
getFieldsValues() {
this.fields = [ { type:'DropdownField',value:'1' }, { type:'Autocompleter',value:'2' } ];
let values = [];
for (let field of this.fields) {
values.push(provider.getFieldValue(field));
}
return values;
}
}
Now when you want to register new field types you can register a provider for them in the factory and don't have to modify your field code.
new Field().getFieldsValues();
I'm writing a tiny reactive framework where I need to find out which subscriber needs updating. I'm implementing deep binding and I'm running into a wall how to find subscribers in an effective manner.
A stored variable can be an object, so for example
{
"user": {
"preferences": {
"food": "vegetarian"
}
}
}
You can get content to any level of this variable like this
getVar("user_preferences_food");
getVar("user_preferences");
However, you can also update it like that
setVar("user_preferences_food", "meat");
setVar("user_preferences", {"food": "meat"});
But in case of the first setVar (user_preferences_food) how can I find the subscriber using getVar("user_preferences"); or even getVar("user"); most effectively.
I already got it working by splitting the var on _ and then one by one concatting the next level and merging all the resulting arrays. But this is very resource intensive. Especially if there are a lot of subscribers. There must be a better way to find them that is less resource intensive.
Edit: I left out part of the explanation.
There is a subscribe method too
subscribe("user", cb);
subscribe("user_preferences", cb);
subscribe("user_preferences_food", cb);
These subscriptions are stored in an array in the framework.
As soon as "user_preferences_food" is updated for example, all subscriptions above should be triggered. But obviously not subscribe('othervar');
simplification of the subscribe method:
var subscriptions = [];
function subscribe(var, callback){
subscriptions.push({var: var, cb: callback});
}
Simplification of getVar
vars = {};
getVar(var){
// find var in vars with this logic: https://stackoverflow.com/a/18937118/249710
// current exact match on subscribers, but need the "parents, grandparents etc here
var toUpdate = _.where(subscriptions, {
"var" : var
});
_.each(toUpdate, function(sub){ sub.cb();});
}
Storing or getting data as part of the key I've already got covered. It is just finding the subscribers in the most effective manner
ps: this is in an environment where I cannot rely on ES6 yet (not all users have it enabled), there is no DOM but I do have underscore included. (Titanium app development platform)
I would try to make a list for the callbacks, so you loop trough one list so you dont have to search, because you know the list is there with all the callbacks.
So if you call setVar('user_prefs') you set a seperate list with the root var. in this case its the user.
if any object is changed with setVar (in depth or not) you go to you're root var, get the list and loop trough this list with the callbacks.
The beauty of this is you can set a list with the root var,
var cbList[FIRSTVAR] this contains all the callbacks. No searching just loop.
Its the mongoDb principle, the data is ready to go, you don't search because you know the list is already there.
You could split the string and use it for reduceing the object.
function getVar(object, path) {
return path
.split('_')
.reduce(function (o, k) {
return (o || {})[k];
}, object);
}
function setVar(object, path, value) {
var keys = path.split('_'),
last = keys.pop();
keys.reduce(function (o, k) {
return o[k] = o[k] || {};
}, object)[last] = value;
}
var object = { user: { preferences: { food: "vegetarian" } } };
console.log(getVar(object, "user_preferences_food"));
console.log(getVar(object, "user_preferences"));
setVar(object, "user_preferences_food", "meat");
console.log(object);
setVar(object, "user_preferences", {"food": "meat"});
console.log(object);
.as-console-wrapper { max-height: 100% !important; top: 0; }
I ended up doing this:
var options = [];
var parts = key.split('_');
var string = parts[0];
_.each(parts, function(p, i){
if (i > 0) string += '_' + p;
options.push(string);
});
var toUpdate = _.filter(subscribers, function(sub){
if (sub.var.indexOf(key + '_') === 0) return true;
if (options.indexOf(sub.var) > -1) return true;
return false;
});
So checking with indexOf on the string to see if there are children. And building an array with parents so any layer is a match, and doing an indexOf on that as well. I think this is the least complicated method of implementing it
Person = Backbone.Model.extend({
defaults: {
name: 'Fetus',
age: 0,
children: []
},
initialize: function(){
alert("Welcome to this world");
},
adopt: function( newChildsName ){
var children_array = this.get("children");
children_array.push( newChildsName );
this.set({ children: children_array });
}
});
var person = new Person({ name: "Thomas", age: 67, children: ['Ryan']});
person.adopt('John Resig');
var children = person.get("children"); // ['Ryan', 'John Resig']
In this example code we have:
children_array = this.get("children")
I was thinking this would just point to the same array in memory (and so would be O(1)). However then I thought that would be a design floor because one could manipulate the array without using this.set() and then event listeners wouldn't fire.
So I'm guessing it (somehow magically) copies the array??
http://backbonejs.org/#Model-set
What happens?
edit: I just found the implementation in the backbone source code at https://github.com/documentcloud/backbone/blob/master/backbone.js (I've pasted relevant code at bottom)
Get returns:
return this.attributes[attr]
so this would just point to the same array in memory right? So one could change the array without using set() and that would be bad.. ? am i correct?
get: function(attr) {
return this.attributes[attr];
},
// Get the HTML-escaped value of an attribute.
escape: function(attr) {
var html;
if (html = this._escapedAttributes[attr]) return html;
var val = this.get(attr);
return this._escapedAttributes[attr] = _.escape(val == null ? '' : '' + val);
},
// Returns `true` if the attribute contains a value that is not null
// or undefined.
has: function(attr) {
return this.get(attr) != null;
},
// Set a hash of model attributes on the object, firing `"change"` unless
// you choose to silence it.
set: function(key, value, options) {
var attrs, attr, val;
// Handle both `"key", value` and `{key: value}` -style arguments.
if (_.isObject(key) || key == null) {
attrs = key;
options = value;
} else {
attrs = {};
attrs[key] = value;
}
// Extract attributes and options.
options || (options = {});
if (!attrs) return this;
if (attrs instanceof Model) attrs = attrs.attributes;
if (options.unset) for (attr in attrs) attrs[attr] = void 0;
// Run validation.
if (!this._validate(attrs, options)) return false;
// Check for changes of `id`.
if (this.idAttribute in attrs) this.id = attrs[this.idAttribute];
var changes = options.changes = {};
var now = this.attributes;
var escaped = this._escapedAttributes;
var prev = this._previousAttributes || {};
// For each `set` attribute...
for (attr in attrs) {
val = attrs[attr];
// If the new and current value differ, record the change.
if (!_.isEqual(now[attr], val) || (options.unset && _.has(now, attr))) {
delete escaped[attr];
(options.silent ? this._silent : changes)[attr] = true;
}
// Update or delete the current value.
options.unset ? delete now[attr] : now[attr] = val;
// If the new and previous value differ, record the change. If not,
// then remove changes for this attribute.
if (!_.isEqual(prev[attr], val) || (_.has(now, attr) !== _.has(prev, attr))) {
this.changed[attr] = val;
if (!options.silent) this._pending[attr] = true;
} else {
delete this.changed[attr];
delete this._pending[attr];
}
}
// Fire the `"change"` events.
if (!options.silent) this.change(options);
return this;
},
The documented interface doesn't actually specify who owns the array reference so you're on your own here. If you look at the implementation, you'll see (as you did) that get just returns a reference straight out of the model's internal attributes. This works fine with immutable types (such as numbers, strings, and booleans) but runs into problems with mutable types such as arrays: you can easily change something without Backbone having any way of knowing about it.
Backbone models appear to be intended to contain primitive types.
There are three reasons to call set:
That's what the interface specification says to do.
If you don't call set, you don't trigger events.
If you don't call set, you'll bypass the validation logic that set has.
You just have to be careful if you're working with array and object values.
Note that this behavior of get and set is an implementation detail and future versions might get smarter about how they handle non-primitive attribute values.
The situation with array attributes (and object attributes for that matter) is actually worse than you might initially suspect. When you say m.set(p, v), Backbone won't consider that set to be a change if v === current_value_of_p so if you pull out an array:
var a = m.get(p);
then modify it:
a.push(x);
and send it back in:
m.set(p, a);
you won't get a "change" event from the model because a === a; Backbone actually uses Underscore's isEqual combined with !== but the effect is the same in this case.
For example, this simple bit of chicanery:
var M = Backbone.Model.extend({});
var m = new M({ p: [ 1 ] });
m.on('change', function() { console.log('changed') });
console.log('Set to new array');
m.set('p', [2]);
console.log('Change without set');
m.get('p').push(3);
console.log('Get array, change, and re-set it');
var a = m.get('p'); a.push(4); m.set('p', a);
console.log('Get array, clone it, change it, set it');
a = _(m.get('p')).clone(); a.push(5); m.set('p', a);
produces two "change" events: one after the first set and one after the last set.
Demo: http://jsfiddle.net/ambiguous/QwZDv/
If you look at set you'll notice that there is some special handling for attributes that are Backbone.Models.
The basic lesson here is simple:
If you're going to use mutable types as attribute values, _.clone them on the way out (or use $.extend(true, ...) if you need a deep copy) if there is any chance that you'll change the value.
A node.js module of mine got too big, so I split it into several smaller (sub)modules.
I copy & pasted all relevant objects into each of the submodules, which now look like
var SOME_CONSTANT = 10;
function my_func() { etc... };
Now I want to export everything in each submodule, en masse, without having to explicitly say exports.SOME_CONSTANT = SOME_CONSTANT a million times (I find that both ugly and error prone).
What is the best way to achieve this?
I assume you do not want to export every local variable.
I will get around to automating this one of these days, but for now I often use this technique.
var x1 = { shouldExport: true } ;
// create a macro in your favorite editor to search and replace so that
x1.name = value ; // instead of var name = value
and
name becomes x1.name
// main body of module
for ( var i in x1) { exports.better_longer_name[i] = x1[i] ;}
//or if you want to add all directly to the export scope
for ( var i in x1) { exports[i] = x1[i] ; }
module.exports = {
SOME_CONSTANT_0 : SOME_CONSTANT_1 ,
SOME_CONSTANT_1 : SOME_CONSTANT_2 ,
SOME_CONSTANT_2 : SOME_CONSTANT_3
}
so why you need that "million" constant to exports?
I'm using Javascript with jQuery. I'd like to implement out params. In C#, it would look something like this:
/*
* odp the object to test
* error a string that will be filled with the error message if odp is illegal. Undefined otherwise.
*
* Returns true if odp is legal.
*/
bool isLegal(odp, out error);
What is the best way to do something like this in JS? Objects?
function isLegal(odp, errorObj)
{
// ...
errorObj.val = "ODP failed test foo";
return false;
}
Firebug tells me that the above approach would work, but is there a better way?
The callback approach mentioned by #Felix Kling is probably the best idea, but I've also found that sometimes it's easy to leverage Javascript object literal syntax and just have your function return an object on error:
function mightFail(param) {
// ...
return didThisFail ? { error: true, msg: "Did not work" } : realResult;
}
then when you call the function:
var result = mightFail("something");
if (result.error) alert("It failed: " + result.msg);
Not fancy and hardly bulletproof, but certainly it's OK for some simple situations.
I think this is pretty much the only way (but I am not a hardcore JavaScript programmer ;)).
What you could also consider is to use a callback function:
function onError(data) {
// do stuff
}
function isLegal(odp, cb) {
//...
if(error) cb(error);
return false;
}
isLegal(value, onError);
Yes, as you yourself mentioned, objects are the best and only way to pass data by reference in JavaScript. I would keep your isLegal function as such and simply call it like this:
var error = {};
isLegal("something", error);
alert(error.val);
The answers I have seen so far aren't implementing out parameters in JavaScript, as they are used in C# (the out keyword). They are merely a workaround that returns an object in case of an error.
But what do you do if you really need out parameters?
Because Javascript doesn't directly support it, you need to build something that is close to C#'s out parameters. Take a look at this approach, I am emulating C#s DateTime.TryParse function in JavaScript. The out parameter is result, and because JavaScript doesn't provide an out keyword, I am using .value inside the function to pass the value outside the function (as inspired by MDN suggestion):
// create a function similar to C#'s DateTime.TryParse
var DateTime = [];
DateTime.TryParse = function(str, result) {
result.value = new Date(str); // out value
return (result.value != "Invalid Date");
};
// now invoke it
var result = [];
if (DateTime.TryParse("05.01.2018", result)) {
alert(result.value);
} else {
alert("no date");
};
Run the snippet and you'll see it works: It parses the str parameter into a Date and returns it in the result parameter. Note that result needs to be initialized as empty array [], before you call the function (it can also be an object{} depending on your needs). This is required because inside the function you "inject" the .value property.
Now you can use the pattern above to write a function as the one in your question (this also shows you how to emulate the new discard parameter known as out _ in C#: In JavaScript we're passing [] as shown below):
// create a function similar to C#'s DateTime.TryParse
var DateTime = [];
DateTime.TryParse = function(str, result) {
result.value = new Date(str); // out value
return (result.value != "Invalid Date");
};
// returns false, if odb is no date, otherwise true
function isLegal(odp, errorObj) {
if (DateTime.TryParse(odp, [])) { // discard result here by passing []
// all OK: leave errorObj.value undefined and return true
return true;
} else {
errorObj.value = "ODP failed test foo"; // return error
return false;
}
}
// now test the function
var odp = "xxx01.12.2018xx"; // invalid date
var errorObj = [];
if (!isLegal(odp, errorObj)) alert(errorObj.value); else alert("OK!");
What this example does is it uses the result parameter to pass an error message as follows:
errorObj.value = "ODP failed test foo"; // return error
If you run the example it will display this message in a popup dialog.
Note: Instead of using a discard parameter as shown above, in JavaScript you could also use a check for undefined, i.e. inside the function check for
if (result === undefined) {
// do the check without passing back a value, i.e. just return true or false
};
Then it is possible to omit result as a parameter completely if not needed, so you could invoke it like
if (DateTime.TryParse(odp)) {
// ... same code as in the snippet above ...
};
I am using a callback method (similar to Felix Kling's approach) to simulate the behavior of out parameters. My answer differs from Kling's in that the callback function acts as a reference-capturing closure rather than a handler.
This approach suffers from JavaScript's verbose anonymous function syntax, but closely reproduces out parameter semantics from other languages.
function isLegal(odp, out_error) {
//...
out_error("ODP failed test foo"); // Assign to out parameter.
return false;
}
var error;
var success = isLegal(null, function (e) { error = e; });
// Invariant: error === "ODP failed test foo".
there is another way JS can pass 'out' parameters. but i believe the best ones for your situation were already mentioned.
Arrays are also passed by reference, not value. thus just as you can pass an object to a function, and then set a property of the object in the function, and then return, and access that object's property, you can similarly pass an Array to a function, set some values of the array inside the function, and return and access those values outside the array.
so in each situation you can ask yourself, "is an array or an object better?"
I'm not going to post any code but what fails to be done here in these answers is to put rhyme to reason. I'm working in the native JS arena and the problem arose that some native API calls need to be transformed because we can't write to the parameters without ugly shameful hacks.
This is my solution:
// Functions that return parameter data should be modified to return
// an array whose zeroeth member is the return value, all other values
// are their respective 1-based parameter index.
That doesn't mean define and return every parameter. Only the
parameters that recieve output.
The reason for this approach is thus: Multiple return values may be needed for any number of procedures. This creates a situation where objects with named values (that ultimately will not be in sync with the lexical context of all operations), constantly need to be memorized in order to appropriately work with the procedure(s).
Using the prescribed method, you only have to know what you called, and where you should be looking rather than having to know what you are looking for.
There is also the advantage that "robust and stupid" alogrithms can be written to wrap around the desired procedure calls to make this operation "more transparent".
It would be wise to use an object, function, or an array (all of which are objects) as a "write-back-output" parameter, but I believe that if any extraneous work must be done, it should be done by the one writing the toolkit to make things easier, or broaden functionality.
This is a one for all answer for every occaision, that keeps APIs looking the way the should at first look, rather than appearing to be and having every resemblence of a hobble-cobbled weave of spaghetti code tapestry that cannot figure out if it is a definition or data.
Congratulations, and good luck.
I'm using the webkitgtk3 and interfaceing some native C Library procs. so this proven code sample might at least serve the purpose of illustration.
// ssize_t read(int filedes, void *buf, size_t nbyte)
SeedValue libc_native_io_read (SeedContext ctx, SeedObject function, SeedObject this_object, gsize argument_count, const SeedValue arguments[], SeedException *exception) {
// NOTE: caller is completely responsible for buffering!
/* C CODING LOOK AND FEEL */
if (argument_count != 3) {
seed_make_exception (ctx, exception, xXx_native_params_invalid,
"read expects 3 arguments: filedes, buffer, nbyte: see `man 3 read' for details",
argument_count
); return seed_make_undefined (ctx);
}
gint filedes = seed_value_to_int(ctx, arguments[0], exception);
void *buf = seed_value_to_string(ctx, arguments[1], exception);
size_t nbyte = seed_value_to_ulong(ctx, arguments[2], exception);
SeedValue result[3];
result[0] = seed_value_from_long(ctx, read(filedes, buf, nbyte), exception);
result[2] = seed_value_from_binary_string(ctx, buf, nbyte, exception);
g_free(buf);
return seed_make_array(ctx, result, 3, exception);
}
The following is approach i am using. And this is answer for this question. However code has not been tested.
function mineCoords( an_x1, an_y1 ) {
this.x1 = an_x1;
this.y1 = an_y1;
}
function mineTest( an_in_param1, an_in_param2 ) {
// local variables
var lo1 = an_in_param1;
var lo2 = an_in_param2;
// process here lo1 and lo2 and
// store result in lo1, lo2
// set result object
var lo_result = new mineCoords( lo1, lo2 );
return lo_result;
}
var lo_test = mineTest( 16.7, 22.4 );
alert( 'x1 = ' + lo_test.x1.toString() + ', y1 = ' + lo_test.y1.toString() );
The usual approach to the specific use case you outlined in Javascript, and in fact most high level languages, is to rely on Errors (aka exceptions) to let you know when something out of the ordinary has occurred. There's no way to pass a value type (strings, numbers etc) by reference in Javascript.
I would just do that. If you really need to feed custom data back to the calling function you can subclass Error.
var MyError = function (message, some_other_param)
{
this.message = message;
this.some_other_param = some_other_param;
}
//I don't think you even need to do this, but it makes it nice and official
MyError.prototype = Error;
...
if (something_is_wrong)
throw new MyError('It failed', /* here's a number I made up */ 150);
Catching exceptions is a pain, I know, but then again so is keeping track of references.
If you really really need something that approaches the behavior of out variables, objects are passed by reference by default, and can handily capture data from other scopes--
function use_out (outvar)
{
outvar.message = 'This is my failure';
return false;
}
var container = { message : '' };
var result = use_out(container );
console.log(container.message); ///gives the string above
console.log(result); //false
I think this goes a some ways towards answering your question, but I think your entire approach is broken from the start. Javascript supports so many much more elegant and powerful ways to get multiple values out of a function. Do some reading about generators, closures, hell even callbacks can be nice in certain situations-- look up continuation passing style.
My point with this whole rant is to encourage anyone reading this to adapt their programming style to the limitations and capabilities of the language they're using, rather than trying to force what they learned from other languages into it.
(BTW some people strongly recommend against closures because they cause evil side-effects, but I wouldn't listen to them. They're purists. Side effects are almost unavoidable in a lot of applications without a lot of tedious backtracking and stepping around cant-get-there-from-here obstacles. If you need them, keeping them all together in a neat lexical scope rather than scattered across a hellscape of obscure pointers and references sounds a lot better to me)
The main advantage of real output parameters is direct modification of one or more scalar variables in the scope of the caller. Among the approaches proposed in other answers, only callbacks satisfy this requirement:
function tryparse_int_1(s, cb)
{ var res = parseInt(s);
cb(res);
return !isNaN( res );
}
function test_1(s)
{ var /* inreger */ i;
if( tryparse_int_1( s, x=>i=x ) )
console.log(`String "${s}" is parsed as integer ${i}.`); else
console.log(`String "${s}" does not start with an integer.`);
}
test_1("47");
test_1("forty-seven");
In this case, passing each output parameter requires five extra characters to wrap its identifier into an anonymous setter function. It is neither very readable nor easy to type frequently, so one can resort to the single most interesting property of scripting languages—their ability to do magick, such as executing strings as code.
The following example implements an extended version of the integer-parsing function above, which now has two output parameters: the resulting integer and a flag indicating whether it is positive:
/* ------------ General emulator of output parameters ------------ */
function out_lit(v)
{ var res;
if( typeof(v) === "string" )
res = '"' + v.split('\"').join('\\\"') + '"'; else
res = `${v}`;
return res;
}
function out_setpar(col, name, value)
{ if( col.outs == undefined ) col.outs = [];
col.outs[name] = value;
}
function out_setret(col, value)
{ col.ret = value; }
function out_ret( col )
{ var s;
for(e in col.outs)
{ s = s + "," + e + "=" + out_lit( col.outs[e] ); }
if( col.ret != undefined )
{ s = s + "," + out_lit( col.ret ); }
return s;
}
/* -------- An intger-parsing function using the emulator -------- */
function tryparse_int_2 // parse the prefix of a string as an integer
( /* string */ s, // in: input string
/* integer */ int, // out: parsed integer value
/* boolean */ pos // out: whether the result is positive
)
{ var /* integer */ res; // function result
var /* array */ col; // collection of out parameters
res = parseInt(s);
col = [];
out_setpar( col, int, res );
out_setpar( col, pos, res > 0 );
out_setret( col, !isNaN( res ) );
return out_ret( col );
}
In this version, passing each output parameters requires two extra characters around its identifier to embed it into a string literal, plus six characters per invocation to evaluate the result:
function test_2(s)
{ var /* integer */ int;
var /* boolean */ pos;
if( !eval( tryparse_int_2( s, "int", "pos" ) ) )
{ console.log(`String "${s}" does not start with an integer.`); }
else
{ if( pos ) adj = "positive";
else adj = "non-positive";
console.log(`String "${s}" is parsed as a ${adj} integer ${int}.`);
}
}
test_2( "55 parrots" );
test_2( "-7 thoughts" );
test_2( "several balls" );
The output of the test code above is:
String "55 parrots" is parsed as a positive integer 55.
String "-7 thoughts" is parsed as a non-positive integer -7.
String "several balls" does not start with an integer.
This solution, however, has a deficiency: it cannot handle returns of non-basic types.
Perhaps a cleaner approach is the emulation of pointers:
// Returns JavaScript for the defintion of a "pointer" to a variable named `v':
// The identifier of the pointer is that of the variable prepended by a $.
function makeref(v)
{ return `var $${v} = {set _(val){${v}=val;},get _() {return ${v};}}`; }
// Calcualtes the square root of `value` and puts it into `$root`.
// Returns whether the operation has succeeded.
// In case of an error, stores error message in `$errmsg`.
function sqrt2
( /* in number */ value, /* value to take the root of */
/* out number */ $root , /* "pointer" to result */
/* out string */ $errmsg /* "pointer" to error message */
)
{ if( typeof( value ) !== "number" )
{ $errmsg._ = "value is not a number.";
return false;
}
if( value < 0 )
{ $errmsg._ = "value is negative.";
return false;
}
$root._ = Math.sqrt(value);
return true;
}
The following test code:
function test(v)
{ var /* string */ resmsg;
var /* number */ root ; eval( makeref( "root" ) );
var /* string */ errmsg; eval( makeref( "errmsg" ) );
if( sqrt2(v, $root, $errmsg) ) resmsg = `Success: ${root}`;
else resmsg = `Error: ${errmsg}`;
console.log(`Square root of ${v}: ` + resmsg );
}
test("s" );
test(-5 );
test( 1.44);
prints:
Square root of s: Error: value is not a number.
Square root of -5: Error: value is negative.
Square root of 1.44: Success: 1.2
"Pointers" created by this method are reusable in other functions and subsequent invocations of the same function. For example, you could define a function that appends strings:
// Append string `sep' to a string pointed to by $s, using `sep` as separator:
// $s shall not point to an undefined value.
function append($s, sep, val)
{ if( $s._ != '' ) $s._ += sep;
$s._ += val;
}
and use it thus:
const sep = ", "
var s; eval( makeref("s") );
s = '';
append( $s, sep, "one" );
append( $s, sep, "two" );
append( $s, sep, "three" );
console.log( s );
It will print:
one, two, three