Related
The function takes an input path like a.b.c and should output a nested structure json like:
{
a: {
b: {
c: {}
}
}
}
The algorithm using iterative style is:
function stringToObj(path, obj) {
var parts = path.split(".");
var part;
while ((part = parts.shift())) {
if (typeof obj[part] != "object") obj[part] = {};
obj = obj[part]; // line 6
}
}
Current usage:
let result = {};
stringToObj("a.b.c", result);
console.log(result); // outputs the json
JsFiddle
The problem:
It relies mutating the obj parameter on line 6.
I would like to not rely upon passing the result object, and rather create one inside the function. Doing so results in different results. A desired example usage:
const result = stringToObj("a.b.c"); // result should be the json
Context:
The exercise is for learning purpose. Main objective is understanding why removing obj and rewriting the function as per follows doesn't work as expected:
function stringToObj(path) {
var obj = {};
var parts = path.split(".");
var part;
while ((part = parts.shift())) {
if (typeof obj[part] != "object") obj[part] = {};
obj = obj[part]; // line 6
}
return obj;
}
After splitting by .s, you can use reduceRight to start at the last property, c, while taking an initial value of an empty object. Inside the callback, use a computed property to return a new object containing the old object at the property being iterated over:
const stringToObj = str => str.split('.').reduceRight(
(lastObj, prop) => ({ [prop]: lastObj }), {}
);
console.log(stringToObj('a.b.c'));
If you're not familiar with it, reduceRight is like reduce, except that it iterates starting from the last element in the array and going backwards, instead of starting from the first element of the array and going forwards. On each iteration, the callback is called, where the first argument (here, lastObj) is the value returned from the last iteration, and the second argument is the current item of the array being iterated over.
You can also reverse the array of properties and use reduce instead of reduceRight, which might be easier to understand at a glance, but it's a bit less elegant:
const stringToObj = str => str.split('.').reverse().reduce(
(lastObj, prop) => ({ [prop]: lastObj }), {}
);
console.log(stringToObj('a.b.c'));
Also, don't mix var and let. If you're going to use ES2015 syntax - which you should - consider always using const, and only use let when you must reassign. Never use var, it has too many gotchas to be worth using in modern code.
I have a bunch of object attributes coming in as dot-delimited strings like "availability_meta.supplier.price", and I need to assign a corresponding value to record['availability_meta']['supplier']['price'] and so on.
Not everything is 3 levels deep: many are only 1 level deep and many are deeper than 3 levels.
Is there a good way to assign this programmatically in Javascript? For example, I need:
["foo.bar.baz", 1] // --> record.foo.bar.baz = 1
["qux.qaz", "abc"] // --> record.qux.qaz = "abc"
["foshizzle", 200] // --> record.foshizzle = 200
I imagine I could hack something together, but I don't have any good algorithm in mind so would appreciate suggestions. I'm using lodash if that's helpful, and open to other libraries that may make quick work of this.
EDIT this is on the backend and run infrequently, so not necessary to optimize for size, speed, etc. In fact code readability would be a plus here for future devs.
EDIT 2 This is NOT the same as the referenced duplicate. Namely, I need to be able to do this assignment multiple times for the same object, and the "duplicate" answer will simply overwrite sub-keys each time. Please reopen!
You mentioned lodash in your question, so I thought I should add their easy object set() and get() functions. Just do something like:
_.set(record, 'availability_meta.supplier.price', 99);
You can read more about it here: https://lodash.com/docs#set
These functions let you do more complex things too, like specify array indexes, etc :)
Something to get you started:
function assignProperty(obj, path, value) {
var props = path.split(".")
, i = 0
, prop;
for(; i < props.length - 1; i++) {
prop = props[i];
obj = obj[prop];
}
obj[props[i]] = value;
}
Assuming:
var arr = ["foo.bar.baz", 1];
You'd call it using:
assignProperty(record, arr[0], arr[1]);
Example: http://jsfiddle.net/x49g5w8L/
What about this?
function convertDotPathToNestedObject(path, value) {
const [last, ...paths] = path.split('.').reverse();
return paths.reduce((acc, el) => ({ [el]: acc }), { [last]: value });
}
convertDotPathToNestedObject('foo.bar.x', 'FooBar')
// { foo: { bar: { x: 'FooBar' } } }
Just do
record['foo.bar.baz'] = 99;
But how would this work? It's strictly for the adventurous with a V8 environment (Chrome or Node harmony), using Object.observe. We observe the the object and capture the addition of new properties. When the "property" foo.bar.baz is added (via an assignment), we detect that this is a dotted property, and transform it into an assignment to record['foo']['bar.baz'] (creating record['foo'] if it does not exist), which in turn is transformed into an assignment to record['foo']['bar']['baz']. It goes like this:
function enable_dot_assignments(changes) {
// Iterate over changes
changes.forEach(function(change) {
// Deconstruct change record.
var object = change.object;
var type = change.type;
var name = change.name;
// Handle only 'add' type changes
if (type !== 'add') return;
// Break the property into segments, and get first one.
var segments = name.split('.');
var first_segment = segments.shift();
// Skip non-dotted property.
if (!segments.length) return;
// If the property doesn't exist, create it as object.
if (!(first_segment in object)) object[first_segment] = {};
var subobject = object[first_segment];
// Ensure subobject also enables dot assignments.
Object.observe(subobject, enable_dot_assignments);
// Set value on subobject using remainder of dot path.
subobject[segments.join('.')] = object[name];
// Make subobject assignments synchronous.
Object.deliverChangeRecords(enable_dot_assignments);
// We don't need the 'a.b' property on the object.
delete object[name];
});
}
Now you can just do
Object.observe(record, enable_dot_assignments);
record['foo.bar.baz'] = 99;
Beware, however, that such assignments will be asynchronous, which may or may not work for you. To solve this, call Object.deliverChangeRecords immediately after the assignment. Or, although not as syntactically pleasing, you could write a helper function, also setting up the observer:
function dot_assignment(object, path, value) {
Object.observe(object, enable_dot_assignments);
object[path] = value;
Object.deliverChangeRecords(enable_dot_assignments);
}
dot_assignment(record, 'foo.bar.baz', 99);
Something like this example perhaps. It will extend a supplied object or create one if it no object is supplied. It is destructive in nature, if you supply keys that already exist in the object, but you can change that if that is not what you want. Uses ECMA5.
/*global console */
/*members split, pop, reduce, trim, forEach, log, stringify */
(function () {
'use strict';
function isObject(arg) {
return arg && typeof arg === 'object';
}
function convertExtend(arr, obj) {
if (!isObject(obj)) {
obj = {};
}
var str = arr[0],
last = obj,
props,
valProp;
if (typeof str === 'string') {
props = str.split('.');
valProp = props.pop();
props.reduce(function (nest, prop) {
prop = prop.trim();
last = nest[prop];
if (!isObject(last)) {
nest[prop] = last = {};
}
return last;
}, obj);
last[valProp] = arr[1];
}
return obj;
}
var x = ['fum'],
y = [
['foo.bar.baz', 1],
['foo.bar.fum', new Date()],
['qux.qaz', 'abc'],
['foshizzle', 200]
],
z = ['qux.qux', null],
record = convertExtend(x);
y.forEach(function (yi) {
convertExtend(yi, record);
});
convertExtend(z, record);
document.body.textContent = JSON.stringify(record, function (key, value, Undefined) {
/*jslint unparam:true */
/*jshint unused:false */
if (value === Undefined) {
value = String(value);
}
return value;
});
}());
it's an old question, but if anyone still looking for a solution can try this
function restructureObject(object){
let result = {};
for(let key in object){
const splittedKeys = key.split('.');
if(splittedKeys.length === 1){
result[key] = object[key];
}
else if(splittedKeys.length > 2){
result = {...result, ...{[splittedKeys.splice(0,1)]: {}} ,...restructureObject({[splittedKeys.join('.')]: object[key]})}
}else{
result[splittedKeys[0]] = {[splittedKeys[1]]: object[key]}
}
}
return result
}
I get from the server a list of objects
[{name:'test01', age:10},{name:'test02', age:20},{name:'test03', age:30}]
I load them into html controls for the user to edit.
Then there is a button to bulk save the entire list back to the database.
Instead of sending the whole list I only want to send the subset of objects that were changed.
It can be any number of items in the array. I want to do something similar to frameworks like Angular that mark an object property like "pristine" when no change has been done to it. Then use that flag to only post to the server the items that are not "pristine", the ones that were modified.
Here is a function down below that will return an array/object of changed objects when supplied with an old array/object of objects and a new array of objects:
// intended to compare objects of identical shape; ideally static.
//
// any top-level key with a primitive value which exists in `previous` but not
// in `current` returns `undefined` while vice versa yields a diff.
//
// in general, the input type determines the output type. that is if `previous`
// and `current` are objects then an object is returned. if arrays then an array
// is returned, etc.
const getChanges = (previous, current) => {
if (isPrimitive(previous) && isPrimitive(current)) {
if (previous === current) {
return "";
}
return current;
}
if (isObject(previous) && isObject(current)) {
const diff = getChanges(Object.entries(previous), Object.entries(current));
return diff.reduce((merged, [key, value]) => {
return {
...merged,
[key]: value
}
}, {});
}
const changes = [];
if (JSON.stringify(previous) === JSON.stringify(current)) {
return changes;
}
for (let i = 0; i < current.length; i++) {
const item = current[i];
if (JSON.stringify(item) !== JSON.stringify(previous[i])) {
changes.push(item);
}
}
return changes;
};
For Example:
const arr1 = [1, 2, 3, 4]
const arr2 = [4, 4, 2, 4]
console.log(getChanges(arr1, arr2)) // [4,4,2]
const obj1 = {
foo: "bar",
baz: [
1, 2, 3
],
qux: {
hello: "world"
},
bingo: "name-o",
}
const obj2 = {
foo: "barx",
baz: [
1, 2, 3, 4
],
qux: {
hello: null
},
bingo: "name-o",
}
console.log(getChanges(obj1.foo, obj2.foo)) // barx
console.log(getChanges(obj1.bingo, obj2.bingo)) // ""
console.log(getChanges(obj1.baz, obj2.baz)) // [4]
console.log(getChanges(obj1, obj2)) // {foo:'barx',baz:[1,2,3,4],qux:{hello:null}}
const obj3 = [{ name: 'test01', age: 10 }, { name: 'test02', age: 20 }, { name: 'test03', age: 30 }]
const obj4 = [{ name: 'test01', age: 10 }, { name: 'test02', age: 20 }, { name: 'test03', age: 20 }]
console.log(getChanges(obj3, obj4)) // [{name:'test03', age:20}]
Utility functions used:
// not required for this example but aid readability of the main function
const typeOf = o => Object.prototype.toString.call(o);
const isObject = o => o !== null && !Array.isArray(o) && typeOf(o).split(" ")[1].slice(0, -1) === "Object";
const isPrimitive = o => {
switch (typeof o) {
case "object": {
return false;
}
case "function": {
return false;
}
default: {
return true;
}
}
};
You would simply have to export the full list of edited values client side, compare it with the old list, and then send the list of changes off to the server.
Hope this helps!
Here are a few ideas.
Use a framework. You spoke of Angular.
Use Proxies, though Internet Explorer has no support for it.
Instead of using classic properties, maybe use Object.defineProperty's set/get to achieve some kind of change tracking.
Use getter/setting functions to store data instead of properties: getName() and setName() for example. Though this the older way of doing what defineProperty now does.
Whenever you bind your data to your form elements, set a special property that indicates if the property has changed. Something like __hasChanged. Set to true if any property on the object changes.
The old school bruteforce way: keep your original list of data that came from the server, deep copy it into another list, bind your form controls to the new list, then when the user clicks submit, compare the objects in the original list to the objects in the new list, plucking out the changed ones as you go. Probably the easiest, but not necessarily the cleanest.
A different take on #6: Attach a special property to each object that always returns the original version of the object:
var myData = [{name: "Larry", age: 47}];
var dataWithCopyOfSelf = myData.map(function(data) {
Object.assign({}, data, { original: data });
});
// now bind your form to dataWithCopyOfSelf.
Of course, this solution assumes a few things: (1) that your objects are flat and simple since Object.assign() doesn't deep copy, (2) that your original data set will never be changed, and (3) that nothing ever touches the contents of original.
There are a multitude of solutions out there.
With ES6 we can use Proxy
to accomplish this task: intercept an Object write, and mark it as dirty.
Proxy allows to create a handler Object that can trap, manipulate, and than forward changes to the original target Object, basically allowing to reconfigure its behavior.
The trap we're going to adopt to intercept Object writes is the handler set().
At this point we can add a non-enumerable property flag like i.e: _isDirty using Object.defineProperty() to mark our Object as modified, dirty.
When using traps (in our case the handler's set()) no changes are applied nor reflected to the Objects, therefore we need to forward the argument values to the target Object using Reflect.set().
Finally, to retrieve the modified objects, filter() the Array with our proxy Objects in search of those having its own Property "_isDirty".
// From server:
const dataOrg = [
{id:1, name:'a', age:10},
{id:2, name:'b', age:20},
{id:3, name:'c', age:30}
];
// Mirror data from server to observable Proxies:
const data = dataOrg.map(ob => new Proxy(ob, {
set() {
Object.defineProperty(ob, "_isDirty", {value: true}); // Flag
return Reflect.set(...arguments); // Forward trapped args to ob
}
}));
// From now on, use proxied data. Let's change some values:
data[0].name = "Lorem";
data[0].age = 42;
data[2].age = 31;
// Collect modified data
const dataMod = data.filter(ob => ob.hasOwnProperty("_isDirty"));
// Test what we're about to send back to server:
console.log(JSON.stringify(dataMod, null, 2));
Without using .defineProperty()
If for some reason you don't feel comfortable into tapping into the original object adding extra properties as flags, you could instead populate immediately
the dataMod (array with modified Objects) with references:
const dataOrg = [
{id:1, name:'a', age:10},
{id:2, name:'b', age:20},
{id:3, name:'c', age:30}
];
// Prepare array to hold references to the modified Objects
const dataMod = [];
const data = dataOrg.map(ob => new Proxy(ob, {
set() {
if (dataMod.indexOf(ob) < 0) dataMod.push(ob); // Push reference
return Reflect.set(...arguments);
}
}));
data[0].name = "Lorem";
data[0].age = 42;
data[2].age = 31;
console.log(JSON.stringify(dataMod, null, 2));
Can I Use - Proxy (IE)
Proxy - handler.set()
Global Objects - Reflect
Reflect.set()
Object.defineProperty()
Object.hasOwnProperty()
Without having to get fancy with prototype properties you could simply store them in another array whenever your form control element detects a change
Something along the lines of:
var modified = [];
data.forEach(function(item){
var domNode = // whatever you use to match data to form control element
domNode.addEventListener('input',function(){
if(modified.indexOf(item) === -1){
modified.push(item);
}
});
});
Then send the modified array to server when it's time to save
Why not use Ember.js observable properties ? You can use the Ember.observer function to get and set changes in your data.
Ember.Object.extend({
valueObserver: Ember.observer('value', function(sender, key, value, rev) {
// Executes whenever the "value" property changes
// See the addObserver method for more information about the callback arguments
})
});
The Ember.object actually does a lot of heavy lifting for you.
Once you define your object, add an observer like so:
object.addObserver('propertyKey', targetObject, targetAction)
My idea is to sort object keys and convert object to be string to compare:
// use this function to sort keys, and save key=>value in an array
function objectSerilize(obj) {
let keys = Object.keys(obj)
let results = []
keys.sort((a, b) => a > b ? -1 : a < b ? 1 : 0)
keys.forEach(key => {
let value = obj[key]
if (typeof value === 'object') {
value = objectSerilize(value)
}
results.push({
key,
value,
})
})
return results
}
// use this function to compare
function compareObject(a, b) {
let aStr = JSON.stringify(objectSerilize(a))
let bStr = JSON.stringify(objectSerilize(b))
return aStr === bStr
}
This is what I think up.
It would be cleanest, I’d think to have the object emit an event when a property is added or removed or modified.
A simplistic implementation could involve an array with the object keys; whenever a setter or heck the constructor returns this, it first calls a static function returning a promise; resolving: map with changed values in the array: things added, things removed, or neither. So one could get(‘changed’) or so forth; returning an array.
Similarly every setter can emit an event with arguments for initial value and new value.
Assuming classes are used, you could easily have a static method in a parent generic class that can be called through its constructor and so really you could simplify most of this by passing the object either to itself, or to the parent through super(checkMeProperty).
Based on my understanding of the docs (here and here) one would need a reference to the memory address for it to work:
const foo = {};
const map = new Map();
map.set(foo,'123'); // Can only be done if memory address of `foo` is known. Any other shimming would require stringification of foo
This is because JavaScript object {} keys can only be strings (at least in ES5).
Yet I see Map shim being available : https://github.com/zloirock/core-js#map. I tried reading the source but its too neatly abstracted (internally uses strong collection which then imports 10 more files)
Question
Answer any of the following please
Is there a simple trick to it and can it truly even be done (without stringification)?
Perhaps it mutates foo to store some string on it and then uses that as the key?
Something else and maybe I am reading the docs wrong?
There are two ways that come to mind. First, obviously, you can have an array of keys, and search it linearly:
Map1 = {
keys: [],
values: [],
};
Map1.set = function(key, val) {
var k = this.keys.indexOf(key);
if(k < 0)
this.keys[k = this.keys.length] = key;
this.values[k] = val;
};
Map1.get = function(key) {
return this.values[this.keys.indexOf(key)];
};
foo = {};
bar = {};
Map1.set(foo, 'xxx');
Map1.set(bar, 'yyy');
document.write(Map1.get(foo) + Map1.get(bar) + "<br>")
The second option is to add a special "key" marker to an object which is used as a key:
Map2 = {
uid: 0,
values: {}
};
Map2.set = function(key, val) {
key = typeof key === 'object'
? (key.__uid = key.__uid || ++this.uid)
: String(key);
this.values[key] = val;
};
Map2.get = function(key) {
key = typeof key === 'object'
? key.__uid
: String(key);
return this.values[key];
};
foo = {};
bar = {};
Map2.set(foo, 'xxx');
Map2.set(bar, 'yyy');
document.write(Map2.get(foo) + Map2.get(bar) + "<br>")
Unlike the 1st option, the second one is O(1). It can be done more accurately by making uid non-writable/enumerable. Also, each Map should have its own "uid" name (this can be easily set up in the Map constructor).
The trick is to store in an array and perform the lookup in O(n) time by iterating and using strict comparison—instead of using a true hash function which would be O(1) lookup. For example consider this:
var myObj = {};
var someArray = [{}, {}, myObj, {}];
console.log(someArray.indexOf(myObj)); // returns 2
Here is my implementation from another answer: Javascript HashTable use Object key
function Map() {
var keys = [], values = [];
return {
put: function (key, value) {
var index = keys.indexOf(key);
if(index == -1) {
keys.push(key);
values.push(value);
}
else {
values[index] = value;
}
},
get: function (key) {
return values[keys.indexOf(key)];
}
};
}
Have a look at my polyfill here. I am not advertising my polyfill, rather all I am saying is that it is the simplest and most straightforward I have yet to find, and thus it is the most suitable for learning and educational analysis. Basically, how it works is it uses a lookup table for the keys and a corresponding value table as visualized below.
var k = {}, j = [], m = document, z = NaN;
var m = new Map([
[k, "foobar"], [j, -0xf], [m, true], [z, function(){}]
]);
Index Key Value
##### ################ ################
0. k ({}) "foobar"
1. j ([]) -15
2. m (Document) true
3. z (NaN) function(){}
Internally, each item is stored at a different index, or at least that is the way I like to do it. This is also similar to the way the browser implements it internally. Unfortunately, I have seen some other polyfills that attempt to instead store the key on the object itself, and mess with all the internal methods to hide it, resulting in the entire webpage running 10000% slower and the maps being so slow that it takes nearly a full millisecond just to set and get new properties. Plus, I cannot fathom how many countless hours they waisted just trying to monkey-patch all the internal methods such as hasOwnProperty.
As for how and why my polyfill works, javascript objects are stored at a different place in memory. That is why [] !== [] and indexOf on an array of javascript objects works properly. It is because they are not the same array.
TL;DR version: I want to avoid adding duplicate Javascript objects to an array of similar objects, some of which might be really big. What's the best approach?
I have an application where I'm loading large amounts of JSON data into a Javascript data structure. While it's a bit more complex than this, assume that I'm loading JSON into an array of Javascript objects from a server through a series of AJAX requests, something like:
var myObjects = [];
function processObject(o) {
myObjects.push(o);
}
for (var x=0; x<1000; x++) {
$.getJSON('/new_object.json', processObject);
}
To complicate matters, the JSON:
is in an unknown schema
is of arbitrary length (probably not enormous, but could be in the 100-200 kb range)
might contain duplicates across different requests
My initial thought is to have an additional object to store a hash of each object (via JSON.stringify?) and check against it on each load, like this:
var myHashMap = {};
function processObject(o) {
var hash = JSON.stringify(o);
// is it in the hashmap?
if (!(myHashMap[hash])) {
myObjects.push(o);
// set the hashmap key for future checks
myHashMap[hash] = true;
}
// else ignore this object
}
but I'm worried about having property names in myHashMap that might be 200 kb in length. So my questions are:
Is there a better approach for this problem than the hashmap idea?
If not, is there a better way to make a hash function for a JSON object of arbitrary length and schema than JSON.stringify?
What are the possible issues with super-long property names in an object?
I'd suggest you create an MD5 hash of the JSON.stringify(o) and store that in your hashmap with a reference to your stored object as the data for the hash. And to make sure that there are no object key order differences in the JSON.stringify(), you have to create a copy of the object that orders the keys.
Then, when each new object comes in, you check it against the hash map. If you find a match in the hash map, then you compare the incoming object with the actual object that you've stored to see if they are truly duplicates (since there can be MD5 hash collisions). That way, you have a manageable hash table (with only MD5 hashes in it).
Here's code to create a canonical string representation of an object (including nested objects or objects within arrays) that handles object keys that might be in a different order if you just called JSON.stringify().
// Code to do a canonical JSON.stringify() that puts object properties
// in a consistent order
// Does not allow circular references (child containing reference to parent)
JSON.stringifyCanonical = function(obj) {
// compatible with either browser or node.js
var Set = typeof window === "object" ? window.Set : global.Set;
// poor man's Set polyfill
if (typeof Set !== "function") {
Set = function(s) {
if (s) {
this.data = s.data.slice();
} else {
this.data = [];
}
};
Set.prototype = {
add: function(item) {
this.data.push(item);
},
has: function(item) {
return this.data.indexOf(item) !== -1;
}
};
}
function orderKeys(obj, parents) {
if (typeof obj !== "object") {
throw new Error("orderKeys() expects object type");
}
var set = new Set(parents);
if (set.has(obj)) {
throw new Error("circular object in stringifyCanonical()");
}
set.add(obj);
var tempObj, item, i;
if (Array.isArray(obj)) {
// no need to re-order an array
// but need to check it for embedded objects that need to be ordered
tempObj = [];
for (i = 0; i < obj.length; i++) {
item = obj[i];
if (typeof item === "object") {
tempObj[i] = orderKeys(item, set);
} else {
tempObj[i] = item;
}
}
} else {
tempObj = {};
// get keys, sort them and build new object
Object.keys(obj).sort().forEach(function(item) {
if (typeof obj[item] === "object") {
tempObj[item] = orderKeys(obj[item], set);
} else {
tempObj[item] = obj[item];
}
});
}
return tempObj;
}
return JSON.stringify(orderKeys(obj));
}
And, the algorithm
var myHashMap = {};
function processObject(o) {
var stringifiedCandidate = JSON.stringifyCanonical(o);
var hash = CreateMD5(stringifiedCandidate);
var list = [], found = false;
// is it in the hashmap?
if (!myHashMap[hash] {
// not in the hash table, so it's a unique object
myObjects.push(o);
list.push(myObjects.length - 1); // put a reference to the object with this hash value in the list
myHashMap[hash] = list; // store the list in the hash table for future comparisons
} else {
// the hash does exist in the hash table, check for an exact object match to see if it's really a duplicate
list = myHashMap[hash]; // get the list of other object indexes with this hash value
// loop through the list
for (var i = 0; i < list.length; i++) {
if (stringifiedCandidate === JSON.stringifyCanonical(myObjects[list[i]])) {
found = true; // found an exact object match
break;
}
}
// if not found, it's not an exact duplicate, even though there was a hash match
if (!found) {
myObjects.push(o);
myHashMap[hash].push(myObjects.length - 1);
}
}
}
Test case for jsonStringifyCanonical() is here: https://jsfiddle.net/jfriend00/zfrtpqcL/
Maybe. For example if You know what kind object goes by You could write better indexing and searching system than JS objects' keys. But You could only do that with JavaScript and object keys are written in C...
Must Your hashing be lossless or not? If can than try to lose compression (MD5). I guessing You will lose some speed and gain some memory. By the way, do JSON.stringify(o) guarantees same key ordering. Because {foo: 1, bar: 2} and {bar: 2, foo: 1} is equal as objects, but not as strings.
Cost memory
One possible optimization:
Instead of using getJSON use $.get and pass "text" as dataType param. Than You can use result as Your hash and convert to object afterwards.
Actually by writing last sentence I though about another solution:
Collect all results with $.get into array
Sort it with buildin (c speed) Array.sort
Now You can easily spot and remove duplicates with one for
Again different JSON strings can make same JavaScript object.