We Keep Coding

Find path in tree by direct path

I have similiary problem like here: JavaScript: Find all parents for element in tree recursive
But I don't find path by name but by direct path.
const path = ["name1", "name4", "name5"];
const data = [
{
'name': 'name1',
'tree': [
{'name': 'name2'},
{'name': 'name3'},
{
'name': 'name4',
'tree': [
{'name': 'name5'},
{'name': 'name6'}
]
},
{'name': 'name7'}
]
},
{
'name': 'name8',
'tree': [
{'name': 'name9'}
]
}
];
It returns every possible path or nothing.
When path is too short, it returns nothing.
When path is too long, it returns nothing.
Thanks for help!
Examples of desired output:
const path = ["name1", "name4", "name5"];
findAPath(data, path)
Returns: ["name1", "name4", "name5"]
const path = ["name1", "name7", "name5"];
findAPath(data, path)
Returns []
const path = ["name1", "name4", "name5", "name5"];
findAPath(data, path)
Returns []
My trying:
let index = 0;
function find(data, index) {
let index = index;
data.some((o) => {
if(o.name == path[index]) {
index++;
find(o.tree, index);
}
});
// I don't know what return here.
// I need to probably return path where I am.
return <>;
}

using Array.prototype.flatMap
Here's a functional solution using a mutual recursion technique -
const None =
Symbol ()
const findPath = (tree = [], names = [], r = []) =>
tree.length && names.length // base: and
? tree.flatMap(branch => findPath1(branch, names, r))
: tree.length || names.length // inductive: xor
? []
: [ r ] // inductive: nor // inductive: nor
const findPath1 = ({ name = "", tree = [] } = {}, [ q = None, ...more ] = [], r = []) =>
name === "" && q === None // base: and
? [ r ]
: name === "" || q === None || name !== q // inductive: xor
? []
: findPath(tree, more, [ ...r, q ]) // inductive: nor
findPath(data, ["name1", "name4", "name5"])
// => [ [ "name1", "name4", "name5" ] ]
NB if your data contains multiple paths to the input values, all paths will be returned -
const data = [
{
'name': 'name1', // name1
'tree': [
{'name': 'name2'},
{'name': 'name3'},
{
'name': 'name4', // name1->name4
'tree': [
{'name': 'name5'}, // name1->name4->name5
{'name': 'name6'}
]
},
{
'name': 'name4', // name1->name4
'tree': [
{'name': 'name5'}, // name1->name4->name5
{'name': 'name6'}
]
},
{'name': 'name7'}
]
},
{
'name': 'name8',
'tree': [
{'name': 'name9'}
]
}
]
Just like you asked, it returns every possible path, or nothing -
findPath(data, ["name1", "name4", "name5"])
// => [ [ "name1", "name4", "name5" ],
// [ "name1", "name4", "name5" ] ]
findPath(data, [ "name1", "name7" ])
// => [ [ "name1", "name7" ] ]
findPath(data, [ "name1", "name9" ])
// => []
When a path is too short or too long, it will return nothing -
findPath(data, [ "name1", "name4" ])
// => []
findPath(data, [ "name1", "name4", "name5", "name6" ])
// => []
Expand the snippet below to verify the results in your own browser -
const None =
Symbol ()
const findPath = (tree = [], names = [], r = []) =>
tree.length && names.length
? tree.flatMap(branch => findPath1(branch, names, r))
: tree.length || names.length
? []
: [ r ]
const findPath1 = ({ name = "", tree = [] } = {}, [ q = None, ...more ] = [], r = []) =>
name === "" && q === None
? [ r ]
: name === "" || q === None || name !== q
? []
: findPath(tree, more, [ ...r, q ])
const data = [
{
'name': 'name1',
'tree': [
{'name': 'name2'},
{'name': 'name3'},
{
'name': 'name4',
'tree': [
{'name': 'name5'},
{'name': 'name6'}
]
},
{'name': 'name7'}
]
},
{
'name': 'name8',
'tree': [
{'name': 'name9'}
]
}
]
console.log(findPath(data, ["name1", "name4", "name5"]))
// [ [ "name1", "name4", "name5" ] ]
console.log(findPath(data, [ "name1", "name7" ]))
// [ [ "name1", "name7" ] ]
console.log(findPath(data, [ "name1", "name9" ]))
// []
using Generators
Here's an alternative implementation using generators -
const None =
Symbol ()
const findPath = function* (tree = [], names = [], r = [])
{ if (tree.length && names.length) // base: and
for (const branch of tree)
yield* findPath1(branch, names, r)
else if (tree.length || names.length) // inductive: xor
return
else // inductive: nor
yield r
}
const findPath1 = function* ({ name = "", tree = [] } = {}, [ q = None, ...more ] = [], r = [])
{ if (name === "" && q === None) // base: and
yield r
else if (name === "" || q === None || name !== q) // inductive: xor
return
else // inductive: nor
yield* findPath(tree, more, [ ...r, q ])
}
It has the exact same output as above, only to coerce the iterable generator into an array, we use Array.from -
Array.from(findPath(data, ["name1", "name4", "name5"]))
// => [ [ "name1", "name4", "name5" ] ]
Array.from(findPath(data, [ "name1", "name7" ]))
// => [ [ "name1", "name7" ] ]
Array.from(findPath(data, [ "name1", "name9" ]))
// => []
Expand the snippet below to verify the results in your own browser -
const None =
Symbol ()
const findPath = function* (tree = [], names = [], r = [])
{ if (tree.length && names.length)
for (const branch of tree)
yield* findPath1(branch, names, r)
else if (tree.length || names.length)
return
else
yield r
}
const findPath1 = function* ({ name = "", tree = [] } = {}, [ q = None, ...more ] = [], r = [])
{ if (name === "" && q === None)
yield r
else if (name === "" || q === None || name !== q)
return
else
yield* findPath(tree, more, [ ...r, q ])
}
const data = [
{
'name': 'name1',
'tree': [
{'name': 'name2'},
{'name': 'name3'},
{
'name': 'name4',
'tree': [
{'name': 'name5'},
{'name': 'name6'}
]
},
{'name': 'name7'}
]
},
{
'name': 'name8',
'tree': [
{'name': 'name9'}
]
}
]
console.log(Array.from(findPath(data, ["name1", "name4", "name5"])))
// [ [ "name1", "name4", "name5" ] ]
console.log(Array.from(findPath(data, [ "name1", "name7" ])))
// [ [ "name1", "name7" ] ]
console.log(Array.from(findPath(data, [ "name1", "name9" ])))
// []
how they're the same; how they're not
Note the similarity between the two implementations and how the result is formed. Both use mutual recursion. The functional solution uses expressions whereas the generator solution uses statements. The generator implementation extends a distinct advantage where by we can chose to stop or continue iteration ("finding") whenever we want.
For example, imagine an input where there are ten (10) unique paths for the given input. Perhaps we want to just return the first match,
const findFirst = (tree = [], names = []) =>
{ for (const path of findPath(tree, names))
return path
}
Or get the first three (3) matches -
const findFirst3 = (tree = [], names = []) =>
{ const r = []
for (const path of findPath(tree, names))
if (r.length < 3)
r.push(path)
return r
}
Or get the first N -
const findFirstN = (tree = [], names = [], n = 0) =>
{ const r = []
for (const path of findPath(tree, names))
if (r.length < n)
r.push(path)
return r
}
Generators are flexible like this. By contrast, the flatMap implementation is eager and always returns all results.

Duplicating an array's elements using functional programming

I'm trying to duplicate each element in an array, but using functional style.
I have this currently:
["a", "b", "c"]
And I'm getting this:
["a","a","b","b","c","c"]
So far I have tried the following, mapping each element to an array, then using flat() to get a 1d array. Is there a cleaner way because it feels like I'm abusing map and flat.
["a", "b", "c"].map(item => [item, item]).flat();
Is there a better way to do this?
I was trying to provide a example as simple as possible but left some details out. The real input is not sorted because elements are not comparable.
It's something like:
[
{
a:"a"
b:"b"
},
{
c: 1
d: 2
},
{
apple: {},
sellers: ["me", "her"]
}
]
The duplicated result should be something like this, where duplicated elements are next to each other:
[
{
a:"a"
b:"b"
},
{
a:"a"
b:"b"
},
{
c: 1
d: 2
},
{
c: 1
d: 2
},
{
apple: {},
sellers: ["me", "her"]
},
{
apple: {},
sellers: ["me", "her"]
}
]

Array.reduce is semantically the appropriate method here: take an object (in this case an array) and return an object of a different type, or with a different length or shape (note: edited to use Array.push for faster performance per #slider suggestion):
EDIT: I've edited my answer to reflect OP's updated input data. Note also, that this solution is cross-browser and NodeJS compatible without requiring transpilation.
let data = [
{
a:"a",
b:"b",
},
{
c: 1,
d: 2
},
{
apple: {},
sellers: ["me", "her"]
}
];
let result = data
.reduce((acc, el) => {
acc.push(el, el);
return acc;
}, []);
console.log(JSON.stringify(result, null, 2));
Otherwise you could map each element, duplicating it, then combine them:
let data = [
{
a:"a",
b:"b",
},
{
c: 1,
d: 2
},
{
apple: {},
sellers: ["me", "her"]
}
];
let result = data.map(item => [item, item]).reduce((acc, arr) => acc.concat(arr));
console.log(JSON.stringify(result, null, 2));
As mentioned in other answers here, either of these approaches have the advantage of not requiring the original array to have been sorted.

You can use the function reduce and concatenate the same object on each iteration.
let array = ["a", "b", "c"],
result = array.reduce((a, c) => a.concat(c, c), []);
console.log(result);
.as-console-wrapper { max-height: 100% !important; top: 0; }

I would recommend Array.prototype.flatMap -
const twice = x =>
[ x, x ]
console .log
( [ 'a', 'b', 'c' ] .flatMap (twice) // [ 'a', 'a', 'b', 'b', 'c', 'c' ]
, [ 1, 2, 3, 4, 5 ] .flatMap (twice) // [ 1, 1, 2, 2, 3, 3, 4, 4, 5, 5 ]
)
flatMap is useful for all kinds of things -
const tree =
[ 0, [ 1 ], [ 2, [ 3 ], [ 4, [ 5 ] ] ] ]
const all = ([ value, ...children ]) =>
[ value ] .concat (children .flatMap (all))
console .log (all (tree))
// [ 0, 1, 2, 3, 4, 5 ]
really cool things -
const ranks =
[ 'J', 'Q', 'K', 'A' ]
const suits =
[ '♡', '♢', '♤', '♧' ]
console .log
( ranks .flatMap (r =>
suits .flatMap (s =>
[ [ r, s ] ]
)
)
)
// [ ['J','♡'], ['J','♢'], ['J','♤'], ['J','♧']
// , ['Q','♡'], ['Q','♢'], ['Q','♤'], ['Q','♧']
// , ['K','♡'], ['K','♢'], ['K','♤'], ['K','♧']
// , ['A','♡'], ['A','♢'], ['A','♤'], ['A','♧']
// ]
flatMap is just a specialised Array.prototype.reduce and is easy to implement in environments where Array.prototype.flatMap is not already supported -
const identity = x =>
x
const flatMap = (xs = [], f = identity) =>
xs .reduce ((r, x) => r . concat (f (x)), [])
const ranks =
[ 'J', 'Q', 'K', 'A' ]
const suits =
[ '♡', '♢', '♤', '♧' ]
console.log
( flatMap (ranks, r =>
flatMap (suits, s =>
[ [ r, s ] ]
)
)
)
// [ ['J','♡'], ['J','♢'], ['J','♤'], ['J','♧']
// , ['Q','♡'], ['Q','♢'], ['Q','♤'], ['Q','♧']
// , ['K','♡'], ['K','♢'], ['K','♤'], ['K','♧']
// , ['A','♡'], ['A','♢'], ['A','♤'], ['A','♧']
// ]

You could just do this:
var arr = ["a", "b", "c"];
arr = arr.concat(arr).sort();
This is one of the simplest methods to do what you are asking to do.

The simplest solution is to use flatMap():
const source = ["a", "b", "c"];
const result = source.flatMap(item => [item, item]);
[ 'a', 'a', 'b', 'b', 'c', 'c' ]
A little bit of classic:
let source = ["a", "b", "c"];
const originalLength = source.length;
for(let i = 0; i <= originalLength + (originalLength - 2); i++) {
source.splice(i, 0, source[i++]);
}
[ 'a', 'a', 'b', 'b', 'c', 'c' ]

Numeric sort on an array with javascript

I have this array :
[ [ 1, 'a' ], [ 2, 'b' ], [ 1, 'd' ], [ 9, 'e' ], [ 1, 'f' ], [ 11, 'g' ], [ 9, 'h' ], [ 3, 'i' ] ]
and I would like to have :
[ [ 11, 'g' ], [ 9, 'e' ], [ 9, 'h' ], [ 3, 'i' ], [ 2, 'b' ], [ 1, 'a' ], [ 1, 'd' ], [ 1, 'f' ] ]
How can I do that with javascript please ?
I tried sort(), I also tried with sort(compare) with :
function compare(x, y) {
return x - y;
}

You can use .sort() with Array Destructuring like this:
function compare([a], [b]) {
return b - a;
}
Demo:
let a = [ [ 1, 'a' ], [ 2, 'b' ], [ 1, 'd' ], [ 9, 'e' ], [ 1, 'f' ], [ 11, 'g' ], [ 9, 'h' ], [ 3, 'i' ] ];
a.sort(compare);
function compare([a], [b]) {
return b - a;
}
console.log(a);
.as-console-wrapper { max-height: 100% !important; top: 0; }
In case first elements match, you can sort based on 2nd element as well:
function compare([a, c], [b, d]) {
return (b - a) || c.localeCompare(d)
}
Demo:
let a = [ [ 2, 'b' ], [ 1, 'd' ], [ 9, 'e' ], [ 1, 'f' ], [ 11, 'g' ], [ 9, 'h' ], [ 3, 'i' ], [ 1, 'a' ] ];
a.sort(compare);
function compare([a, c], [b, d]) {
return (b - a) || c.localeCompare(d);
}
console.log(a);
.as-console-wrapper { max-height: 100% !important; top: 0 }

You need to compare the first element in the nested array since you want to sort based on that number.
function compare(x, y) {
return y[0] - x[0];
}
var data = [
[1, 'a'],
[2, 'b'],
[1, 'd'],
[9, 'e'],
[1, 'f'],
[11, 'g'],
[9, 'h'],
[3, 'i']
];
function compare(x, y) {
return y[0] - x[0];
}
data.sort(compare);
console.log(data);
In case you want to sort based on second element(secondary sorting in case the first element is same) then use String#localeCompare method for comparing.
function compare(x, y) {
return y[0] - x[0] || x[1].localeCompare(y[0]);
}
var data = [
[2, 'b'],
[1, 'd'],
[9, 'e'],
[1, 'f'],
[1, 'a'],
[11, 'g'],
[9, 'h'],
[3, 'i']
];
function compare(x, y) {
return (y[0] - x[0]) || x[1].localeCompare(y[1]);
}
data.sort(compare);
console.log(data);

Compare the elements based on their first element, which is the number.
var a = [ [ 1, 'a' ], [ 2, 'b' ], [ 1, 'd' ], [ 9, 'e' ], [ 1, 'f' ], [ 11, 'g' ], [ 9, 'h' ], [ 3, 'i' ] ];
a = a.sort((a,b) => {
return b[0] - a[0]
});
console.log(a)

Use sort to compare first element and if first element is same then compare the second element.
arr.sort( (a,b) => (b[0] - a[0]) || (b[1] - a[1]) )
Demo
var arr = [ [ 1, 'a' ], [ 2, 'b' ], [ 1, 'd' ], [ 9, 'e' ], [ 1, 'f' ], [ 11, 'g' ], [ 9, 'h' ], [ 3, 'i' ] ];
arr.sort( (a,b) => (b[0] - a[0]) || (b[1] - a[1]) );
console.log(arr);

Breaking an array of users in a group of sub-arrays

I would like to break a array with users:
[
{name: "Carlos"},
{name: "Marcos"},
{name: "Fernando"},
{name: "Jhon"},
{name: "Loius"},
{name: "Jacob"},
]
And get something like this:
[
[
{name: "Jhon"},
{name: "Loius"},
{name: "Jacob"},
],
[
{name: "Carlos"},
{name: "Marcos"},
{name: "Fernando"},
]
]
The criterion for splitting them is that I want that each sub array to have a maximum of 3 users, but the number of sub arrays can be unlimited.

function splitIntoParts(input, maxElementsPerPart) {
const inputClone = [...input]; // create a copy because splice modifies the original array reference.
const result = [];
const parts = Math.ceil(input.length/maxElementsPerPart);
for(let i = 0; i < parts; i++) {
result.push(inputClone.splice(0, maxElementsPerPart));
}
return result;
}
console.log(splitIntoParts([
{name: "Carlos"},
{name: "Marcos"},
{name: "Fernando"},
{name: "Jhon"},
{name: "Loius"},
{name: "Jacob"},
{name: "Simon"},
], 3));

chunk is elegantly expressed using functional style
const chunk = (xs = [], n = 1) =>
xs.length <= n
? [ xs ]
: [ xs.slice (0, n) ] .concat (chunk (xs.slice (n), n))
const data =
[ 1, 2, 3, 4, 5, 6 ]
console.log (chunk (data, 1))
// [ [ 1 ], [ 2 ], [ 3 ], [ 4 ], [ 5 ], [ 6 ] ]
console.log (chunk (data, 2))
// [ [ 1, 2 ], [ 3, 4 ], [ 5, 6 ] ]
console.log (chunk (data, 3))
// [ [ 1, 2, 3 ], [ 4, 5, 6 ] ]
console.log (chunk (data, 4))
// [ [ 1, 2, 3, 4 ], [ 5, 6 ] ]
console.log (chunk ())
// [ [ ] ]
I think take and drop abstractions make the function read a littler nicer. Your opinion may vary.
const take = (xs = [], n = 1) =>
xs.slice (0, n)
const drop = (xs = [], n = 1) =>
xs.slice (n)
const chunk = (xs = [], n = 1) =>
xs.length <= n
? [ xs ]
: [ take (xs, n) ] .concat (chunk (drop (xs, n), n))

let data = [
{name: "Carlos"},
{name: "Marcos"},
{name: "Fernando"},
{name: "Jhon"},
{name: "Loius"},
{name: "Jacob"},
]
function toGroupsOf(n, array){
return Array.from(
{length: Math.ceil(array.length/n)}, //how many groups
(_,i) => array.slice(i*n, (i+1)*n) //get the items for this group
)
}
console.log(toGroupsOf(3, data));
.as-console-wrapper{top:0;max-height:100%!important}
or
function toGroupsOf(n, array){
var groups = Array(Math.ceil(array.length/n));
for(var i=0; i<groups.length; ++i)
groups[i] = array.slice(i*n, (i+1)*n);
return groups;
}

javascript: Trying to flatten array only one level

I am trying to write a function to flatten an array. I have part of the function working and I need help in the other half.
flatten: function(anyArray, singleLevel) {
if (singleLevel == true) {
flatArray = Array.prototype.concat.apply([], anyArray);
return flatArray;
}
flatArray = Array.prototype.concat.apply([], anyArray);
if (flatArray.length != anyArray.length) {
flatArray = someObject.array.flatten(flatArray);
}
return flatArray;
}
if I type
.flatten([[[1],[1,2,3,[4,5],4],[2,3]]], true);
I want it to flatten only one level:
[[1],[1,2,3,[4,5],4],[2,3]]

Modern JavaScript allows us to handle this very easily using a variety of techniques
Using Array.prototype.flat -
const arr =
[ [ 1 ], [ 2, 3, [ 4, 5, [ 6 ] ] ], [ 7, [ 8, 9 ] ] ]
const flatArr =
arr.flat(1) // 1 is default depth
console.log(JSON.stringify(arr))
console.log(JSON.stringify(flatArr))
// [[1],[2,3,[4,5,[6]]],[7,[8,9]]]
// [1,2,3,[4,5,[6]],7,[8,9]]
Using Array.prototype.flatMap -
const arr =
[ [ 1 ], [ 2, 3, [ 4, 5, [ 6 ] ] ], [ 7, [ 8, 9 ] ] ]
const flatArr =
arr.flatMap(x => x)
console.log(JSON.stringify(arr))
console.log(JSON.stringify(flatArr))
// [[1],[2,3,[4,5,[6]]],[7,[8,9]]]
// [1,2,3,[4,5,[6]],7,[8,9]]
Using a spread argument to Array.prototype.concat
const arr =
[ [ 1 ], [ 2, 3, [ 4, 5, [ 6 ] ] ], [ 7, [ 8, 9 ] ] ]
const flatArr =
[].concat(...arr)
console.log(JSON.stringify(arr))
console.log(JSON.stringify(flatArr))
// [[1],[2,3,[4,5,[6]]],[7,[8,9]]]
// [1,2,3,[4,5,[6]],7,[8,9]]
Older version of JavaScript (ECMAScript 5 and below) can use techniques like Function.prototype.apply -
var arr =
[ [ 1 ], [ 2, 3, [ 4, 5, [ 6 ] ] ], [ 7, [ 8, 9 ] ] ]
var flatArr =
Array.prototype.concat.apply([], arr)
console.log(JSON.stringify(arr))
console.log(JSON.stringify(flatArr))
// [[1],[2,3,[4,5,[6]]],[7,[8,9]]]
// [1,2,3,[4,5,[6]],7,[8,9]]
Using Array.prototype.reduce -
var arr =
[ [ 1 ], [ 2, 3, [ 4, 5, [ 6 ] ] ], [ 7, [ 8, 9 ] ] ]
var flatArr =
arr.reduce((r, a) => r.concat(a), [])
console.log(JSON.stringify(arr))
console.log(JSON.stringify(flatArr))
// [[1],[2,3,[4,5,[6]]],[7,[8,9]]]
// [1,2,3,[4,5,[6]],7,[8,9]]
Using a primitive for loop -
var arr =
[ [ 1 ], [ 2, 3, [ 4, 5, [ 6 ] ] ], [ 7, [ 8, 9 ] ] ]
var flatArr =
[]
for (var i = 0; i < arr.length; i = i + 1)
flatArr = flatArr.concat(arr[i])
console.log(JSON.stringify(arr))
console.log(JSON.stringify(flatArr))
// [[1],[2,3,[4,5,[6]]],[7,[8,9]]]
// [1,2,3,[4,5,[6]],7,[8,9]]

The concat array method expects one or more arrays as arguments, whose elements will be appended:
[1].concat([2, 3], [4]) // [1, 2, 3, 4]
So if you are using apply, that will flatten another level:
[].concat.apply([1], [[2], [3]]) // === [1].concat([2], [3])
So you can either use push instead of concat, or call (or just direct invocation) instead of apply to get only a single flattening level.

if you use ES6/ES2015 you can use spread operator. Something like this
console.log(...[[[1],[1,2,3,[4,5],4],[2,3]]])