I'd like to encode an Integer into a String using 4 different Bits ( A, B, C, D):
Therefore I've wrote a simple Int to customBase conversion you can find here:
function messageToCustomBase(message, charset) {
var base = charset.length,
integer = message,
result = ""
do {
var index = integer % base
result = charset[index] + result
integer = parseInt(integer/base)
} while (integer > 0)
return result
}
The code is working quite fine and encoding looks like this:
0 --> "A"
1 --> "B"
2 --> "C"
...
100 --> "BCBA"
...
10000 --> "CBDABAA"
But due to a special program, I am practically forced to use a special encode algorithm which converts the integer to a string which ...
1. should have as small a length as possible
2. uses maximum four different letters (4 different bits -> A, B, C, D)
3. Prevents never having the same letters next to each other
"ABADADA" -> legit
"ABCDAD" -> legit
"BABCA" -> legit
"CDDABC" -> not legit (because of "D" "D")
"BBBACAB" -> not legit (because of "B" "B" "B")
Question: > How can I avoid multiple characters following each other by editing my messageToBase function?
Example: Encoding Integer 42 will give the result "CCC".
Because in my special case "CCC" is not valid it should be encoded different (maybe to e.g. "CBCDCA", ...)
Note: The string must be able to be decoded later on so just adding random fill-characters between repeating characters after the decoding will not be a working solution.
Note: Here you can find a working fiddle for the base en/decoding
Edit: Theoretically we have to add in cases where it comes to repeating letters an additional bit ("E"). Otherwise it will just lead to complications with other Integers!
So I really got no clue how to fix this problem and any help would be very appreciated. :)
Here's a (rather brute-force) solution:
This simply iterates over your existing encoder with higher "bit depth" until it finds a solution without repeated letters. The encoding depth is prepended to the encoded value (otherwise it would be impossible to reverse the calculation, because you wouldn't have any way of knowing whether "BA" means 2 (encoded to depth B) or 3 (encoded to depth C).
This doesn't guarantee the lowest possible "bit depth", but it does keep the strings as short as possible given your encoding method.
Here's a demonstration of encoding and decoding numbers from 1 through 50:
var encode = function(number) {
if (number == 1) {
return "A-A"; // hacky workaround for endless loop in messageToCustomBase
}
var ret = "";
var bits = Math.floor(Math.log2(number) + 1); // minimum depth required for this number
var chars = "ABCDEFGHIJKLMNOPQRSTUVWXYZ";
for (var i = bits; i < 26; i++) {
var bitIdentifier = chars.charAt(i - 1);
var encoded = messageToCustomBase(number, chars.substr(0, i));
if (encoded.match(/(.)\1/)) {
// duplicate letters found, keep looking
} else {
return bitIdentifier + "-" + encoded
}
}
}
var decode = function(m) {
if (m === 'A-A') {
return 1
} // hack again
var chars = "ABCDEFGHIJKLMNOPQRSTUVWXYZ";
var parts = m.split('-');
var charset = chars.substr(0, parts[0].charCodeAt(0) - 64);
return customBaseToMessage(parts[1], charset)
}
// exactly as your original code:
function messageToCustomBase(message, charset) {
var base = charset.length,
integer = message,
result = ""
do {
var index = integer % base
result = charset[index] + result
integer = parseInt(integer / base)
} while (integer > 0)
return result
}
function customBaseToMessage(message, charset) {
var base = charset.length,
result = 0
for (var i = 0; i < message.length; i++) {
var index = charset.indexOf(message[i])
result = result * base + index
}
return result
}
// encode numbers for testing:
for (var i = 1; i < 51; i++) {
console.log(i, encode(i), decode(encode(i)));
}
You can do this by sticking to base 3 after the first loop, where the characters available are the three you didn't use for the previous character.
As a modification of your encoding code, this looks like
function messageToCustomBase(message, charset) {
var base = charset.length,
integer = message,
result = "",
previous_index = null
var index = integer % base
result = charset[index] + result
integer = parseInt(integer/base)
previous_index = index
while (integer > 0) {
var index = integer % (base - 1)
if (index >= previous_index) {
index++
}
result = charset[index] + result
integer = parseInt(integer/(base - 1))
previous_index = index
}
return result
}
This is pretty close to optimal, but not quite. The reason is that the index for the left-most character will never be zero. Basically, it's doing a "correct" conversion to base (3..4), and is refusing to have representations that start with 0. If you're willing to forgo that, you can get a bit shorter on average.
I've looked this up online without much results because it's quite hard to describe in a few words.
Basically, I need to have a function, say puntil which takes the argument string. Basically, the function permutes until the string is equal to the argument.
For example if you run puntil('ab') it should do inside the function:
a
b
c
d
e
f
g
h
i
j
k
l
m
n
o
p
q
r
s
t
u
v
w
x
y
z
aa
ab !! MATCH
Another example, for puntil('abcd') it will do
a
b
c
d
e
f
g
h
i
j
k
l
m
n
o
p
q
r
s
t
u
v
w
x
y
z
aa
ab
ac
ad
ae
af
ag
ah
ai
aj
ak
al
am
an
ao
ap
aq
ar
as
at
au
av
aw
ax
ay
az
... etc etc ..
until it matches abcd.
Basically an infinite permutation until it matches.
Any ideas?
Here is the fiddle
var alphabet = ['a','b','c'];//,'d','e','f','g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v','w','x','y','z'];
var output = "";
var match = "cccc"; //<----------- match here
//This is your main function
function permutate() {
var d = 0; // d for depth
while (true) {
//Your main alphabet iteration
for (var i=0; i<alphabet.length; i++){
//First level
if (d === 0) {
console.log(alphabet[i])
output = alphabet[i];
}
else
iterator(alphabet[i], d); //Call iterator for d > 0
if (output === match)
return;
}
d++; //Increase the depth
}
}
//This function runs n depths of iterations
function iterator(s, depth){
if (depth === 0)
return;
for (var i=0; i<alphabet.length; i++){
if (depth > 1)
iterator(s + alphabet[i], depth - 1)
else {
console.log(s + alphabet[i]);
output = s + alphabet[i];
}
if (output === match)
return;
}
};
Explanation:
Your program needs to traverse a tree of alphabet like this
[a]
-[a]
-[a]
-[a]...
-[b]...
[b] ...
-[b] -
-[a]...
-[b]...
[b] - ...
[c] - ...
This could have easily been done through a conventional recursive function if not for the requirement that you need to finish each depth first.
So we need a special iterator(s, depth) function which can perform number of nested iterations (depth) requested.
So the main function can call the iterator with increasing depths (d++).
That's all!!
Warning: This is a prototype only. This can be optimized and improved. It uses global variables for the ease of demonstrating. Your real program should avoid globals. Also I recommend calling the iterator() inside setTimeout if your match word is too long.
The n depths can only be limited by your resources.
Fiddle here
function next(charArray, rightBound){
if(!rightBound){
rightBound = charArray.length;
}
var oldValue = charArray[rightBound-1];
var newValue = nextCharacter(charArray[rightBound-1]);
charArray[rightBound-1] = newValue;
if(newValue < oldValue){
if(rightBound > 1){
next(charArray, rightBound-1);
}
else{
charArray.push('a');
}
}
return charArray;
}
function nextCharacter(char){
if(char === 'z'){
return 'a'
}
else{
return String.fromCharCode(char.charCodeAt(0) + 1)
}
}
function permuteUntil(word){
var charArray = ['a'];
var wordChain = ['a'];
while(next(charArray).join('') !== word){
wordChain.push(charArray.join(''));
}
wordChain.push(word);
return wordChain.join(', ');
}
alert(permuteUntil('ab'));
What OP is asking is a bit ambiguous, so I'll post for both the things (that I doubt) OP is asking.
First, the question can be, what will be the position of input string in the infinite permutation of alphabets (which I see as more legit question, I've given the reason later). This can be done in the following manner:
Taking an example (input = dhca). So, all strings of 1 to 3 characters length will come before this string. So, add: 26^1 + 26^2 + 26^3 to the answer. Then, 1st character is d, which means, following the dictionary, if 1st character is a | b | c, all characters past that are valid. So, add 3 * 26^3 to the answer. Now, say 1st character is d. Then, we can have all characters from a to g (7) and last 2 characters can be anything. So, add 7 * 26^2 to the answer. Going on in this way, we get the answer as:
26^1 + 26^2 + 26^3 + (3 * 26^3) + (7 * 26^2) + (2 * 26^1) + (0) + 1
= 75791
OK. Now the second thing, which I think OP is actually asking (to print all strings before we get a match). Now, why I think this is unfeasible is because if we have input as zzzzz (5 characters long) we need to print 26^1 + 26^2 + 26^3 + 26^4 + 26^5 strings, which is 12356630. So, for this part, I assume max length of input string is 5 (And definitely no more) because for 6 character length string, we need to print ~321272406 strings --> NOT POSSIBLE.
So, a simple solution to this can be:
Create an array of size 27 as: arr[] = {'', 'a', 'b', ..., 'y', 'z'}. 1st character is null.
Write 5 (max string length) nested loops from 0 to 26 (inclusive) and add it to dummy string and print it. Something like.
for i from 0 to 26
String a = "" + arr[i]
for j from 0 to 26
String b = a + arr[j]
for k from 0 to 26
String c = b + arr[k]
for l from 0 to 26
String d = c + arr[l]
for m from 0 to 26
String e = d + arr[m]
print e
if(e equals input string)
break from all loops //use some flag, goto statement etc.
You asked for a more elegant solution, here's a simple function that converts integers into lowercase strings of characters allowing you to easily iterate through strings.
function toWord(val, first, out) {
if(first == 1)
out = String.fromCharCode(97 + val % 26) + out;
else
out = String.fromCharCode(97 + (val-1) % 26) + out;
if(val % 26 == 0 && first == 0) {
val -= 26;
}
else {
val -= val %26;
}
val = val / 26;
if(val != 0)
return toWord(val, 0, out);
else
return out;
}
It's by no means perfect, but it's short and simple. When calling the function set val to be the integer you want to convert, first as 1, and out as "".
For example the following will apply yourFunction to the first 10,000 lowercase strings
for(int i=0; i<10000; i++) {
youFunction(toWord(i, 1, ""));
}
So you need to always start incrementing from a? Since they are char values you can easily do this with the following construct:
Note that this is a java solution :)
public static char[] increment(char[] arr, int pos) {
// get current position char
char currentChar = arr[pos];
// increment it by one
currentChar++;
// if it is at the end of it's range
boolean overflow = false;
if (currentChar > 'Z') {
overflow = true;
currentChar = 'A';
}
// always update current char
arr[pos] = currentChar;
if (overflow) {
if (pos == 0) {
// resize array and add new character
char[] newArr = new char[arr.length + 1];
System.arraycopy(arr, 0, newArr, 0, arr.length);
newArr[arr.length] = 'A';
arr = newArr;
} else {
// overflowed, increment one position back
arr = increment(arr, pos - 1);
}
}
return arr;
}
public static void main(String[] args) {
final String stringToUse = "A";
final String stringToSearch = "ABCD";
char[] use = stringToUse.toCharArray();
for (;;) {
final String result = new String(use);
System.out.println("Candidate: " + result);
if (stringToSearch.equals(result)) {
System.out.println("Found!");
break;
}
use = increment(use, use.length - 1);
}
}
In the past I have made a function that generates an unique id (number) from a string. Today I discover that it is not as unique as should be. Never saw a problem before with it. Today two different inputs generates the same id (number).
I use the same technique in Delphi, C++, PHP and Javascript to generate the same id's so there is no difference when different languages are involved to a project. For example this can be handy to communicate, for HTML id's, tempfiles etc.
In general, what I do is calculate a CRC16 of a string, add the sum and return it.
For example, these two strings generate the same id (number):
o.uniqueId( 'M:/Mijn Muziek/Various Artists/Revs & ElBee - Tell It To My Heart.mp3' );
o.uniqueId( 'M:/Mijn Muziek/Various Artists/Dwight Yoakam - The Back Of Your Hand.Mp3');
They both generates an id of 224904.
The following example is a javascript example. My question is, how can i avoid (with a little change) that it generates a duplicate? (In case you might wonder what 'o.' means, it is the object where these functions belongs to):
o.getCrc16 = function(s, bSumPos) {
if(typeof s !== 'string' || s.length === 0) {
return 0;
}
var crc = 0xFFFF,
L = s.length,
sum = 0,
x = 0,
j = 0;
for(var i = 0; i < L; i++) {
j = s.charCodeAt(i);
sum += ((i + 1) * j);
x = ((crc >> 8) ^ j) & 0xFF;
x ^= x >> 4;
crc = ((crc << 8) ^ (x << 12) ^ (x << 5) ^ x) & 0xFFFF;
}
return crc + ((bSumPos ? 1 : 0) * sum);
}
o.uniqueId = function(s, bres) {
if(s == undefined || typeof s != 'string') {
if(!o.___uqidc) {
o.___uqidc = 0;
} else {
++o.___uqidc;
}
var od = new Date(),
i = s = od.getTime() + '' + o.___uqidc;
} else {
var i = o.getCrc16(s, true);
}
return((bres) ? 'res:' : '') + (i + (i ? s.length : 0));
};
How can I avoid duplicates with use of a little change to the code?
All right, did allot of testing and come to this. A relative short unique id generated by the following:
o.lz = function(i,c)
{
if( typeof c != 'number' || c <= 0 || (typeof i != 'number' && typeof i != 'string') )
{ return i; }
i+='';
while( i.length < c )
{ i='0'+i; }
return i;
}
o.getHashCode = function(s)
{
var hash=0,c=(typeof s == 'string')?s.length:0,i=0;
while(i<c)
{
hash = ((hash<<5)-hash)+s.charCodeAt(i++);
//hash = hash & hash; // Convert to 32bit integer
}
return ( hash < 0 )?((hash*-1)+0xFFFFFFFF):hash; // convert to unsigned
};
o.uniqueId = function( s, bres )
{
if( s == undefined || typeof s != 'string' )
{
if( !o.___uqidc )
{ o.___uqidc=0; }
else { ++o.___uqidc; }
var od = new Date(),
i = s = od.getTime()+''+o.___uqidc;
}
else { var i = o.getHashCode( s ); }
return ((bres)?'res:':'')+i.toString(32)+'-'+o.lz((s.length*4).toString(16),3);
};
Examples:
o.uniqueId( 'M:/Mijn Muziek/Various Artists/Revs & ElBee - Tell It To My Heart.mp3' );
o.uniqueId( 'M:/Mijn Muziek/Various Artists/Dwight Yoakam - The Back Of Your Hand.Mp3');
Will produce the following id's:
dh8qi9t-114
je38ugg-120
For my purpose it seems to be unique enough, also the extra length adds some more uniqueness. Test it on filesystem with approx 40.000 mp3 files and did not found any collision.
If you think this is not the way to go, please let me know.
You should increase the number of bits created by your hash function. Assuming that your hash function approximately uniform over the space, you can mathematically derive the probability of observing a collision.
This is strongly related to the birthday paradox. In the case of CRC16, where the hash value is 17 bits (though your implementation may have a mistake; I don't see how you obtained 224094 as that is greater than 2^17), you will have a collision probability above 50% when you store more than approximately 2^8 items. In addition, CRC is not really a great hashing function because it's meant for error detection, not uniform hashing.
This table shows mathematical probabilities of collision based on hash length. For example, if you have a 128-bit hash key, you can store up to 10^31 elements before the collision probability increases beyond 10^-15. As a comparison, this probability is lower than that of your hard drive failing, or maybe your computer being zapped by lightning, so a safe number to use.
Just increase your hash length based on the number of strings you are planning to identify, and a pick a collision probability that is acceptable to you.
How do you convert decimal values to their hexadecimal equivalent in JavaScript?
Convert a number to a hexadecimal string with:
hexString = yourNumber.toString(16);
And reverse the process with:
yourNumber = parseInt(hexString, 16);
If you need to handle things like bit fields or 32-bit colors, then you need to deal with signed numbers. The JavaScript function toString(16) will return a negative hexadecimal number which is usually not what you want. This function does some crazy addition to make it a positive number.
function decimalToHexString(number)
{
if (number < 0)
{
number = 0xFFFFFFFF + number + 1;
}
return number.toString(16).toUpperCase();
}
console.log(decimalToHexString(27));
console.log(decimalToHexString(48.6));
The code below will convert the decimal value d to hexadecimal. It also allows you to add padding to the hexadecimal result. So 0 will become 00 by default.
function decimalToHex(d, padding) {
var hex = Number(d).toString(16);
padding = typeof (padding) === "undefined" || padding === null ? padding = 2 : padding;
while (hex.length < padding) {
hex = "0" + hex;
}
return hex;
}
function toHex(d) {
return ("0"+(Number(d).toString(16))).slice(-2).toUpperCase()
}
For completeness, if you want the two's-complement hexadecimal representation of a negative number, you can use the zero-fill-right shift >>> operator. For instance:
> (-1).toString(16)
"-1"
> ((-2)>>>0).toString(16)
"fffffffe"
There is however one limitation: JavaScript bitwise operators treat their operands as a sequence of 32 bits, that is, you get the 32-bits two's complement.
With padding:
function dec2hex(i) {
return (i+0x10000).toString(16).substr(-4).toUpperCase();
}
The accepted answer did not take into account single digit returned hexadecimal codes. This is easily adjusted by:
function numHex(s)
{
var a = s.toString(16);
if ((a.length % 2) > 0) {
a = "0" + a;
}
return a;
}
and
function strHex(s)
{
var a = "";
for (var i=0; i<s.length; i++) {
a = a + numHex(s.charCodeAt(i));
}
return a;
}
I believe the above answers have been posted numerous times by others in one form or another. I wrap these in a toHex() function like so:
function toHex(s)
{
var re = new RegExp(/^\s*(\+|-)?((\d+(\.\d+)?)|(\.\d+))\s*$/);
if (re.test(s)) {
return '#' + strHex( s.toString());
}
else {
return 'A' + strHex(s);
}
}
Note that the numeric regular expression came from 10+ Useful JavaScript Regular Expression Functions to improve your web applications efficiency.
Update: After testing this thing several times I found an error (double quotes in the RegExp), so I fixed that. HOWEVER! After quite a bit of testing and having read the post by almaz - I realized I could not get negative numbers to work.
Further - I did some reading up on this and since all JavaScript numbers are stored as 64 bit words no matter what - I tried modifying the numHex code to get the 64 bit word. But it turns out you can not do that. If you put "3.14159265" AS A NUMBER into a variable - all you will be able to get is the "3", because the fractional portion is only accessible by multiplying the number by ten(IE:10.0) repeatedly. Or to put that another way - the hexadecimal value of 0xF causes the floating point value to be translated into an integer before it is ANDed which removes everything behind the period. Rather than taking the value as a whole (i.e.: 3.14159265) and ANDing the floating point value against the 0xF value.
So the best thing to do, in this case, is to convert the 3.14159265 into a string and then just convert the string. Because of the above, it also makes it easy to convert negative numbers because the minus sign just becomes 0x26 on the front of the value.
So what I did was on determining that the variable contains a number - just convert it to a string and convert the string. This means to everyone that on the server side you will need to unhex the incoming string and then to determine the incoming information is numeric. You can do that easily by just adding a "#" to the front of numbers and "A" to the front of a character string coming back. See the toHex() function.
Have fun!
After another year and a lot of thinking, I decided that the "toHex" function (and I also have a "fromHex" function) really needed to be revamped. The whole question was "How can I do this more efficiently?" I decided that a to/from hexadecimal function should not care if something is a fractional part but at the same time it should ensure that fractional parts are included in the string.
So then the question became, "How do you know you are working with a hexadecimal string?". The answer is simple. Use the standard pre-string information that is already recognized around the world.
In other words - use "0x". So now my toHex function looks to see if that is already there and if it is - it just returns the string that was sent to it. Otherwise, it converts the string, number, whatever. Here is the revised toHex function:
/////////////////////////////////////////////////////////////////////////////
// toHex(). Convert an ASCII string to hexadecimal.
/////////////////////////////////////////////////////////////////////////////
toHex(s)
{
if (s.substr(0,2).toLowerCase() == "0x") {
return s;
}
var l = "0123456789ABCDEF";
var o = "";
if (typeof s != "string") {
s = s.toString();
}
for (var i=0; i<s.length; i++) {
var c = s.charCodeAt(i);
o = o + l.substr((c>>4),1) + l.substr((c & 0x0f),1);
}
return "0x" + o;
}
This is a very fast function that takes into account single digits, floating point numbers, and even checks to see if the person is sending a hex value over to be hexed again. It only uses four function calls and only two of those are in the loop. To un-hex the values you use:
/////////////////////////////////////////////////////////////////////////////
// fromHex(). Convert a hex string to ASCII text.
/////////////////////////////////////////////////////////////////////////////
fromHex(s)
{
var start = 0;
var o = "";
if (s.substr(0,2).toLowerCase() == "0x") {
start = 2;
}
if (typeof s != "string") {
s = s.toString();
}
for (var i=start; i<s.length; i+=2) {
var c = s.substr(i, 2);
o = o + String.fromCharCode(parseInt(c, 16));
}
return o;
}
Like the toHex() function, the fromHex() function first looks for the "0x" and then it translates the incoming information into a string if it isn't already a string. I don't know how it wouldn't be a string - but just in case - I check. The function then goes through, grabbing two characters and translating those in to ASCII characters. If you want it to translate Unicode, you will need to change the loop to going by four(4) characters at a time. But then you also need to ensure that the string is NOT divisible by four. If it is - then it is a standard hexadecimal string. (Remember the string has "0x" on the front of it.)
A simple test script to show that -3.14159265, when converted to a string, is still -3.14159265.
<?php
echo <<<EOD
<html>
<head><title>Test</title>
<script>
var a = -3.14159265;
alert( "A = " + a );
var b = a.toString();
alert( "B = " + b );
</script>
</head>
<body>
</body>
</html>
EOD;
?>
Because of how JavaScript works in respect to the toString() function, all of those problems can be eliminated which before were causing problems. Now all strings and numbers can be converted easily. Further, such things as objects will cause an error to be generated by JavaScript itself. I believe this is about as good as it gets. The only improvement left is for W3C to just include a toHex() and fromHex() function in JavaScript.
Without the loop:
function decimalToHex(d) {
var hex = Number(d).toString(16);
hex = "000000".substr(0, 6 - hex.length) + hex;
return hex;
}
// Or "#000000".substr(0, 7 - hex.length) + hex;
// Or whatever
// *Thanks to MSDN
Also isn't it better not to use loop tests that have to be evaluated?
For example, instead of:
for (var i = 0; i < hex.length; i++){}
have
for (var i = 0, var j = hex.length; i < j; i++){}
Combining some of these good ideas for an RGB-value-to-hexadecimal function (add the # elsewhere for HTML/CSS):
function rgb2hex(r,g,b) {
if (g !== undefined)
return Number(0x1000000 + r*0x10000 + g*0x100 + b).toString(16).substring(1);
else
return Number(0x1000000 + r[0]*0x10000 + r[1]*0x100 + r[2]).toString(16).substring(1);
}
Constrained/padded to a set number of characters:
function decimalToHex(decimal, chars) {
return (decimal + Math.pow(16, chars)).toString(16).slice(-chars).toUpperCase();
}
For anyone interested, here's a JSFiddle comparing most of the answers given to this question.
And here's the method I ended up going with:
function decToHex(dec) {
return (dec + Math.pow(16, 6)).toString(16).substr(-6)
}
Also, bear in mind that if you're looking to convert from decimal to hex for use in CSS as a color data type, you might instead prefer to extract the RGB values from the decimal and use rgb().
For example (JSFiddle):
let c = 4210330 // your color in decimal format
let rgb = [(c & 0xff0000) >> 16, (c & 0x00ff00) >> 8, (c & 0x0000ff)]
// Vanilla JS:
document.getElementById('some-element').style.color = 'rgb(' + rgb + ')'
// jQuery:
$('#some-element').css('color', 'rgb(' + rgb + ')')
This sets #some-element's CSS color property to rgb(64, 62, 154).
var number = 3200;
var hexString = number.toString(16);
The 16 is the radix and there are 16 values in a hexadecimal number :-)
function dec2hex(i)
{
var result = "0000";
if (i >= 0 && i <= 15) { result = "000" + i.toString(16); }
else if (i >= 16 && i <= 255) { result = "00" + i.toString(16); }
else if (i >= 256 && i <= 4095) { result = "0" + i.toString(16); }
else if (i >= 4096 && i <= 65535) { result = i.toString(16); }
return result
}
If you want to convert a number to a hexadecimal representation of an RGBA color value, I've found this to be the most useful combination of several tips from here:
function toHexString(n) {
if(n < 0) {
n = 0xFFFFFFFF + n + 1;
}
return "0x" + ("00000000" + n.toString(16).toUpperCase()).substr(-8);
}
AFAIK comment 57807 is wrong and should be something like:
var hex = Number(d).toString(16);
instead of
var hex = parseInt(d, 16);
function decimalToHex(d, padding) {
var hex = Number(d).toString(16);
padding = typeof (padding) === "undefined" || padding === null ? padding = 2 : padding;
while (hex.length < padding) {
hex = "0" + hex;
}
return hex;
}
And if the number is negative?
Here is my version.
function hexdec (hex_string) {
hex_string=((hex_string.charAt(1)!='X' && hex_string.charAt(1)!='x')?hex_string='0X'+hex_string : hex_string);
hex_string=(hex_string.charAt(2)<8 ? hex_string =hex_string-0x00000000 : hex_string=hex_string-0xFFFFFFFF-1);
return parseInt(hex_string, 10);
}
As the accepted answer states, the easiest way to convert from decimal to hexadecimal is var hex = dec.toString(16). However, you may prefer to add a string conversion, as it ensures that string representations like "12".toString(16) work correctly.
// Avoids a hard-to-track-down bug by returning `c` instead of `12`
(+"12").toString(16);
To reverse the process you may also use the solution below, as it is even shorter.
var dec = +("0x" + hex);
It seems to be slower in Google Chrome and Firefox, but is significantly faster in Opera. See http://jsperf.com/hex-to-dec.
I'm doing conversion to hex string in a pretty large loop, so I tried several techniques in order to find the fastest one. My requirements were to have a fixed-length string as a result, and encode negative values properly (-1 => ff..f).
Simple .toString(16) didn't work for me since I needed negative values to be properly encoded. The following code is the quickest I've tested so far on 1-2 byte values (note that symbols defines the number of output symbols you want to get, that is for 4-byte integer it should be equal to 8):
var hex = ['0', '1', '2', '3', '4', '5', '6', '7', '8', '9', 'a', 'b', 'c', 'd', 'e', 'f'];
function getHexRepresentation(num, symbols) {
var result = '';
while (symbols--) {
result = hex[num & 0xF] + result;
num >>= 4;
}
return result;
}
It performs faster than .toString(16) on 1-2 byte numbers and slower on larger numbers (when symbols >= 6), but still should outperform methods that encode negative values properly.
Converting hex color numbers to hex color strings:
A simple solution with toString and ES6 padStart for converting hex color numbers to hex color strings.
const string = `#${color.toString(16).padStart(6, '0')}`;
For example:
0x000000 will become #000000
0xFFFFFF will become #FFFFFF
Check this example in a fiddle here
How to convert decimal to hexadecimal in JavaScript
I wasn't able to find a brutally clean/simple decimal to hexadecimal conversion that didn't involve a mess of functions and arrays ... so I had to make this for myself.
function DecToHex(decimal) { // Data (decimal)
length = -1; // Base string length
string = ''; // Source 'string'
characters = [ '0', '1', '2', '3', '4', '5', '6', '7', '8', '9', 'A', 'B', 'C', 'D', 'E', 'F' ]; // character array
do { // Grab each nibble in reverse order because JavaScript has no unsigned left shift
string += characters[decimal & 0xF]; // Mask byte, get that character
++length; // Increment to length of string
} while (decimal >>>= 4); // For next character shift right 4 bits, or break on 0
decimal += 'x'; // Convert that 0 into a hex prefix string -> '0x'
do
decimal += string[length];
while (length--); // Flip string forwards, with the prefixed '0x'
return (decimal); // return (hexadecimal);
}
/* Original: */
D = 3678; // Data (decimal)
C = 0xF; // Check
A = D; // Accumulate
B = -1; // Base string length
S = ''; // Source 'string'
H = '0x'; // Destination 'string'
do {
++B;
A& = C;
switch(A) {
case 0xA: A='A'
break;
case 0xB: A='B'
break;
case 0xC: A='C'
break;
case 0xD: A='D'
break;
case 0xE: A='E'
break;
case 0xF: A='F'
break;
A = (A);
}
S += A;
D >>>= 0x04;
A = D;
} while(D)
do
H += S[B];
while (B--)
S = B = A = C = D; // Zero out variables
alert(H); // H: holds hexadecimal equivalent
You can do something like this in ECMAScript 6:
const toHex = num => (num).toString(16).toUpperCase();
If you are looking for converting Large integers i.e. Numbers greater than Number.MAX_SAFE_INTEGER -- 9007199254740991, then you can use the following code
const hugeNumber = "9007199254740991873839" // Make sure its in String
const hexOfHugeNumber = BigInt(hugeNumber).toString(16);
console.log(hexOfHugeNumber)
To sum it all up;
function toHex(i, pad) {
if (typeof(pad) === 'undefined' || pad === null) {
pad = 2;
}
var strToParse = i.toString(16);
while (strToParse.length < pad) {
strToParse = "0" + strToParse;
}
var finalVal = parseInt(strToParse, 16);
if ( finalVal < 0 ) {
finalVal = 0xFFFFFFFF + finalVal + 1;
}
return finalVal;
}
However, if you don't need to convert it back to an integer at the end (i.e. for colors), then just making sure the values aren't negative should suffice.
I haven't found a clear answer, without checks if it is negative or positive, that uses two's complement (negative numbers included). For that, I show my solution to one byte:
((0xFF + number +1) & 0x0FF).toString(16);
You can use this instruction to any number bytes, only you add FF in respective places. For example, to two bytes:
((0xFFFF + number +1) & 0x0FFFF).toString(16);
If you want cast an array integer to string hexadecimal:
s = "";
for(var i = 0; i < arrayNumber.length; ++i) {
s += ((0xFF + arrayNumber[i] +1) & 0x0FF).toString(16);
}
In case you're looking to convert to a 'full' JavaScript or CSS representation, you can use something like:
numToHex = function(num) {
var r=((0xff0000&num)>>16).toString(16),
g=((0x00ff00&num)>>8).toString(16),
b=(0x0000ff&num).toString(16);
if (r.length==1) { r = '0'+r; }
if (g.length==1) { g = '0'+g; }
if (b.length==1) { b = '0'+b; }
return '0x'+r+g+b; // ('#' instead of'0x' for CSS)
};
var dec = 5974678;
console.log( numToHex(dec) ); // 0x5b2a96
This is based on Prestaul and Tod's solutions. However, this is a generalisation that accounts for varying size of a variable (e.g. Parsing signed value from a microcontroller serial log).
function decimalToPaddedHexString(number, bitsize)
{
let byteCount = Math.ceil(bitsize/8);
let maxBinValue = Math.pow(2, bitsize)-1;
/* In node.js this function fails for bitsize above 32bits */
if (bitsize > 32)
throw "number above maximum value";
/* Conversion to unsigned form based on */
if (number < 0)
number = maxBinValue + number + 1;
return "0x"+(number >>> 0).toString(16).toUpperCase().padStart(byteCount*2, '0');
}
Test script:
for (let n = 0 ; n < 64 ; n++ ) {
let s=decimalToPaddedHexString(-1, n);
console.log(`decimalToPaddedHexString(-1,${(n+"").padStart(2)}) = ${s.padStart(10)} = ${("0b"+parseInt(s).toString(2)).padStart(34)}`);
}
Test results:
decimalToPaddedHexString(-1, 0) = 0x0 = 0b0
decimalToPaddedHexString(-1, 1) = 0x01 = 0b1
decimalToPaddedHexString(-1, 2) = 0x03 = 0b11
decimalToPaddedHexString(-1, 3) = 0x07 = 0b111
decimalToPaddedHexString(-1, 4) = 0x0F = 0b1111
decimalToPaddedHexString(-1, 5) = 0x1F = 0b11111
decimalToPaddedHexString(-1, 6) = 0x3F = 0b111111
decimalToPaddedHexString(-1, 7) = 0x7F = 0b1111111
decimalToPaddedHexString(-1, 8) = 0xFF = 0b11111111
decimalToPaddedHexString(-1, 9) = 0x01FF = 0b111111111
decimalToPaddedHexString(-1,10) = 0x03FF = 0b1111111111
decimalToPaddedHexString(-1,11) = 0x07FF = 0b11111111111
decimalToPaddedHexString(-1,12) = 0x0FFF = 0b111111111111
decimalToPaddedHexString(-1,13) = 0x1FFF = 0b1111111111111
decimalToPaddedHexString(-1,14) = 0x3FFF = 0b11111111111111
decimalToPaddedHexString(-1,15) = 0x7FFF = 0b111111111111111
decimalToPaddedHexString(-1,16) = 0xFFFF = 0b1111111111111111
decimalToPaddedHexString(-1,17) = 0x01FFFF = 0b11111111111111111
decimalToPaddedHexString(-1,18) = 0x03FFFF = 0b111111111111111111
decimalToPaddedHexString(-1,19) = 0x07FFFF = 0b1111111111111111111
decimalToPaddedHexString(-1,20) = 0x0FFFFF = 0b11111111111111111111
decimalToPaddedHexString(-1,21) = 0x1FFFFF = 0b111111111111111111111
decimalToPaddedHexString(-1,22) = 0x3FFFFF = 0b1111111111111111111111
decimalToPaddedHexString(-1,23) = 0x7FFFFF = 0b11111111111111111111111
decimalToPaddedHexString(-1,24) = 0xFFFFFF = 0b111111111111111111111111
decimalToPaddedHexString(-1,25) = 0x01FFFFFF = 0b1111111111111111111111111
decimalToPaddedHexString(-1,26) = 0x03FFFFFF = 0b11111111111111111111111111
decimalToPaddedHexString(-1,27) = 0x07FFFFFF = 0b111111111111111111111111111
decimalToPaddedHexString(-1,28) = 0x0FFFFFFF = 0b1111111111111111111111111111
decimalToPaddedHexString(-1,29) = 0x1FFFFFFF = 0b11111111111111111111111111111
decimalToPaddedHexString(-1,30) = 0x3FFFFFFF = 0b111111111111111111111111111111
decimalToPaddedHexString(-1,31) = 0x7FFFFFFF = 0b1111111111111111111111111111111
decimalToPaddedHexString(-1,32) = 0xFFFFFFFF = 0b11111111111111111111111111111111
Thrown: 'number above maximum value'
Note: Not too sure why it fails above 32 bitsize
rgb(255, 255, 255) // returns FFFFFF
rgb(255, 255, 300) // returns FFFFFF
rgb(0,0,0) // returns 000000
rgb(148, 0, 211) // returns 9400D3
function rgb(...values){
return values.reduce((acc, cur) => {
let val = cur >= 255 ? 'ff' : cur <= 0 ? '00' : Number(cur).toString(16);
return acc + (val.length === 1 ? '0'+val : val);
}, '').toUpperCase();
}
Arbitrary precision
This solution take on input decimal string, and return hex string. A decimal fractions are supported. Algorithm
split number to sign (s), integer part (i) and fractional part (f) e.g for -123.75 we have s=true, i=123, f=75
integer part to hex:
if i='0' stop
get modulo: m=i%16 (in arbitrary precision)
convert m to hex digit and put to result string
for next step calc integer part i=i/16 (in arbitrary precision)
fractional part
count fractional digits n
multiply k=f*16 (in arbitrary precision)
split k to right part with n digits and put them to f, and left part with rest of digits and put them to d
convert d to hex and add to result.
finish when number of result fractional digits is enough
// #param decStr - string with non-negative integer
// #param divisor - positive integer
function dec2HexArbitrary(decStr, fracDigits=0) {
// Helper: divide arbitrary precision number by js number
// #param decStr - string with non-negative integer
// #param divisor - positive integer
function arbDivision(decStr, divisor)
{
// algorithm https://www.geeksforgeeks.org/divide-large-number-represented-string/
let ans='';
let idx = 0;
let temp = +decStr[idx];
while (temp < divisor) temp = temp * 10 + +decStr[++idx];
while (decStr.length > idx) {
ans += (temp / divisor)|0 ;
temp = (temp % divisor) * 10 + +decStr[++idx];
}
if (ans.length == 0) return "0";
return ans;
}
// Helper: calc module of arbitrary precision number
// #param decStr - string with non-negative integer
// #param mod - positive integer
function arbMod(decStr, mod) {
// algorithm https://www.geeksforgeeks.org/how-to-compute-mod-of-a-big-number/
let res = 0;
for (let i = 0; i < decStr.length; i++)
res = (res * 10 + +decStr[i]) % mod;
return res;
}
// Helper: multiply arbitrary precision integer by js number
// #param decStr - string with non-negative integer
// #param mult - positive integer
function arbMultiply(decStr, mult) {
let r='';
let m=0;
for (let i = decStr.length-1; i >=0 ; i--) {
let n = m+mult*(+decStr[i]);
r= (i ? n%10 : n) + r
m= n/10|0;
}
return r;
}
// dec2hex algorithm starts here
let h= '0123456789abcdef'; // hex 'alphabet'
let m= decStr.match(/-?(.*?)\.(.*)?/) || decStr.match(/-?(.*)/); // separate sign,integer,ractional
let i= m[1].replace(/^0+/,'').replace(/^$/,'0'); // integer part (without sign and leading zeros)
let f= (m[2]||'0').replace(/0+$/,'').replace(/^$/,'0'); // fractional part (without last zeros)
let s= decStr[0]=='-'; // sign
let r=''; // result
if(i=='0') r='0';
while(i!='0') { // integer part
r=h[arbMod(i,16)]+r;
i=arbDivision(i,16);
}
if(fracDigits) r+=".";
let n = f.length;
for(let j=0; j<fracDigits; j++) { // frac part
let k= arbMultiply(f,16);
f = k.slice(-n);
let d= k.slice(0,k.length-n);
r+= d.length ? h[+d] : '0';
}
return (s?'-':'')+r;
}
// -----------
// TESTS
// -----------
let tests = [
["0",2],
["000",2],
["123",0],
["-123",0],
["00.000",2],
["255.75",5],
["-255.75",5],
["127.999",32],
];
console.log('Input Standard Abitrary');
tests.forEach(t=> {
let nonArb = (+t[0]).toString(16).padEnd(17,' ');
let arb = dec2HexArbitrary(t[0],t[1]);
console.log(t[0].padEnd(10,' '), nonArb, arb);
});
// Long Example (40 digits after dot)
let example = "123456789012345678901234567890.09876543210987654321"
console.log(`\nLong Example:`);
console.log('dec:',example);
console.log('hex: ',dec2HexArbitrary(example,40));
The problem basically how many padding zeros to expect.
If you expect string 01 and 11 from Number 1 and 17. it's better to use Buffer as a bridge, with which number is turn into bytes, and then the hex is just an output format of it. And the bytes organization is well controlled by Buffer functions, like writeUInt32BE, writeInt16LE, etc.
import { Buffer } from 'buffer';
function toHex(n) { // 4byte
const buff = Buffer.alloc(4);
buff.writeInt32BE(n);
return buff.toString('hex');
}
> toHex(1)
'00000001'
> toHex(17)
'00000011'
> toHex(-1)
'ffffffff'
> toHex(-1212)
'fffffb44'
> toHex(1212)
'000004bc'
Here's my solution:
hex = function(number) {
return '0x' + Math.abs(number).toString(16);
}
The question says: "How to convert decimal to hexadecimal in JavaScript". While, the question does not specify that the hexadecimal string should begin with a 0x prefix, anybody who writes code should know that 0x is added to hexadecimal codes to distinguish hexadecimal codes from programmatic identifiers and other numbers (1234 could be hexadecimal, decimal, or even octal).
Therefore, to correctly answer this question, for the purpose of script-writing, you must add the 0x prefix.
The Math.abs(N) function converts negatives to positives, and as a bonus, it doesn't look like somebody ran it through a wood-chipper.
The answer I wanted, would have had a field-width specifier, so we could for example show 8/16/32/64-bit values the way you would see them listed in a hexadecimal editing application. That, is the actual, correct answer.