Develop Reference

document.getElementsByClassName exact match to class

There are two similar classes - 'item' and 'item one'
When I use document.getElementsByClassName('item') it returns all elements that match both classes above.
How I can get elements with 'item' class only?

The classname item one means the element has class item and class one.
So, when you do document.getElementsByClassName('item'), it returns that element too.
You should do something like this to select the elements with only the class item:
e = document.getElementsByClassName('item');
for(var i = 0; i < e.length; i++) {
// Only if there is only single class
if(e[i].className == 'item') {
// Do something with the element e[i]
alert(e[i].className);
}
}
This will check that the elements have only class item.
Live Demo

document.querySelectorAll('.item:not(.one)');
(see querySelectorAll)
The other way is to loop over the what document.getElementsByClassName('item') returns, and check if the one class is present (or not):
if(element.classList.contains('one')){
...
}

You're going to want to make your own function for exact matches, because spaces in a class means it has multiple classes. Something like:
function GetElementsByExactClassName(someclass) {
var i, length, elementlist, data = [];
// Get the list from the browser
elementlist = document.getElementsByClassName(someclass);
if (!elementlist || !(length = elementlist.length))
return [];
// Limit by elements with that specific class name only
for (i = 0; i < length; i++) {
if (elementlist[i].className == someclass)
data.push(elementlist[i]);
}
// Return the result
return data;
} // GetElementsByExactClassName

You can use Array.filter to filter the matched set to be only those with class test:
var elems = Array.filter( document.getElementsByClassName('test'), function(el){
return !!el.className.match(/\s*test\s*/);
});
Or only those with test but not one:
var elems = Array.filter( document.getElementsByClassName('test'), function(el){
return el.className.indexOf('one') < 0;
});
(Array.filter may work differently depending on your browser, and is not available in older browsers.) For best browser compatibility, jQuery would be excellent for this: $('.test:not(.one)')

If you have jQuery, it can be done using the attribute equals selector syntax like this: $('[class="item"]').
If you insist on using document.getElementsByClassName, see the other answers.

jQuery: Getting "secondary" class

I'm using jQuery to loop through some elements in a document. These elements are of type <tr> with class .input-row. Some of the elements can also have a secondary class (for example .input-area).
I use $(".input-row").each() to loop through the elements, but how can I determine if the $(this)-element has a "secondary" class and if so, get the name of this?

$(".input-row").each(function(){
$.each($(this).attr('class').split(/\s+/), function(i, v){
if(v !== 'input-row'){
alert(v);
}
});
});

You can get the list of all classes using attr, and you could simply get rid of the the class you used to select the elements:
$(".input-row").each(function() {
var allClasses = $(this).attr("class"),
otherClasses = allClasses.replace("input-row", "");
});
If the element has more than two classes, then you will be left with a list of the other classes. To separate them, you could split on the spaces.

You can get all the classes on the element and then remove the one you know exists('input-row'), then you will be left with any other classes on that element.
$('.input-row').each(function() {
var classes = $.trim($(this).attr('class').replace('input-row', ''));
if (classes.length > 1) {
alert(classes) // Names of all of the remaining classes
}
}
If you want to get them individually, you can do classes.split(' ') to get an array.

I prefer a declarative approach. You know which classes there can be, so write cases for them:
$(".input-row.input-area").each( /* ... */ );
$(".input-row.input-field").each( /* ... */ );
If you for some reason don't know the class name, you can extract it along the lines of:
$(".input-row").each(function () {
var myClasses = this.className.split(" ");
$.each(myClasses, function (i, className) {
if (className != 'input-row') {
alert(className);
}
});
});

I'd go for this
var otherClasses = $('.input-row').attr('class').split(' ');
otherClasses.splice($.inArray("input-row", otherClasses),1);
Simply doing this
var otherClasses = $('.input-row').attr('class').replace('input-row', '').split(' ');
Will give you an extra string in the array with an empty value in it, because of the space behind "input-row"

You could also do something like this to detect the number of classes:
var nc = $('#myId').attr('class').split(' ').length;
Then if nc is greater than 1:
$('#myID').attr('class').split(' ')[1]
should return the 'secondary' class
See a fiddle

Getting element by a custom attribute using JavaScript

I have an XHTML page where each HTML element has a unique custom attribute, like this:
<div class="logo" tokenid="14"></div>
I need a way to find this element by ID, similar to document.getElementById(), but instead of using a general ID, I want to search for the element using my custom "tokenid" attribute. Something like this:
document.getElementByTokenId('14');
Is that possible? If yes - any hint would be greatly appreciated.
Thanks.

It is not good to use custom attributes in the HTML. If any, you should use HTML5's data attributes.
Nevertheless you can write your own function that traverses the tree, but that will be quite slow compared to getElementById because you cannot make use of any index:
function getElementByAttribute(attr, value, root) {
root = root || document.body;
if(root.hasAttribute(attr) && root.getAttribute(attr) == value) {
return root;
}
var children = root.children,
element;
for(var i = children.length; i--; ) {
element = getElementByAttribute(attr, value, children[i]);
if(element) {
return element;
}
}
return null;
}
In the worst case, this will traverse the whole tree. Think about how to change your concept so that you can make use browser functions as much as possible.
In newer browsers you use of the querySelector method, where it would just be:
var element = document.querySelector('[tokenid="14"]');
This will be much faster too.
Update: Please note #Andy E's comment below. It might be that you run into problems with IE (as always ;)). If you do a lot of element retrieval of this kind, you really should consider using a JavaScript library such as jQuery, as the others mentioned. It hides all these browser differences.

<div data-automation="something">
</div>
document.querySelector("div[data-automation]")
=> finds the div
document.querySelector("div[data-automation='something']")
=> finds the div with a value

If you're using jQuery, you can use some of their selector magic to do something like this:
$('div[tokenid=14]')
as your selector.

You can accomplish this with JQuery:
$('[tokenid=14]')
Here's a fiddle for an example.

If you're willing to use JQuery, then:
var myElement = $('div[tokenid="14"]').get();

Doing this with vanilla JavaScript will do the trick:
const something = document.querySelectorAll('[data-something]')

Use this more stable Function:
function getElementsByAttribute(attr, value) {
var match = [];
/* Get the droids we are looking for*/
var elements = document.getElementsByTagName("*");
/* Loop through all elements */
for (var ii = 0, ln = elements.length; ii < ln; ii++) {
if (elements[ii].nodeType === 1){
if (elements[ii].name != null){
/* If a value was passed, make sure it matches the elements */
if (value) {
if (elements[ii].getAttribute(attr) === value)
match.push(elements[ii]);
} else {
/* Else, simply push it */
match.push(elements[ii]);
}
}
}
}
return match;
};

How to Get Element By Class in JavaScript?

I want to replace the contents within a html element so I'm using the following function for that:
function ReplaceContentInContainer(id,content) {
var container = document.getElementById(id);
container.innerHTML = content;
}
ReplaceContentInContainer('box','This is the replacement text');
<div id='box'></div>
The above works great but the problem is I have more than one html element on a page that I want to replace the contents of. So I can't use ids but classes instead. I have been told that javascript does not support any type of inbuilt get element by class function. So how can the above code be revised to make it work with classes instead of ids?
P.S. I don't want to use jQuery for this.

This code should work in all browsers.
function replaceContentInContainer(matchClass, content) {
var elems = document.getElementsByTagName('*'), i;
for (i in elems) {
if((' ' + elems[i].className + ' ').indexOf(' ' + matchClass + ' ')
> -1) {
elems[i].innerHTML = content;
}
}
}
The way it works is by looping through all of the elements in the document, and searching their class list for matchClass. If a match is found, the contents is replaced.
jsFiddle Example, using Vanilla JS (i.e. no framework)

Of course, all modern browsers now support the following simpler way:
var elements = document.getElementsByClassName('someClass');
but be warned it doesn't work with IE8 or before. See http://caniuse.com/getelementsbyclassname
Also, not all browsers will return a pure NodeList like they're supposed to.
You're probably still better off using your favorite cross-browser library.

document.querySelectorAll(".your_class_name_here");
That will work in "modern" browsers that implement that method (IE8+).
function ReplaceContentInContainer(selector, content) {
var nodeList = document.querySelectorAll(selector);
for (var i = 0, length = nodeList.length; i < length; i++) {
nodeList[i].innerHTML = content;
}
}
ReplaceContentInContainer(".theclass", "HELLO WORLD");
If you want to provide support for older browsers, you could load a stand-alone selector engine like Sizzle (4KB mini+gzip) or Peppy (10K mini) and fall back to it if the native querySelector method is not found.
Is it overkill to load a selector engine just so you can get elements with a certain class? Probably. However, the scripts aren't all that big and you will may find the selector engine useful in many other places in your script.

A Simple and an easy way
var cusid_ele = document.getElementsByClassName('custid');
for (var i = 0; i < cusid_ele.length; ++i) {
var item = cusid_ele[i];
item.innerHTML = 'this is value';
}

I'm surprised there are no answers using Regular Expressions. This is pretty much Andrew's answer, using RegExp.test instead of String.indexOf, since it seems to perform better for multiple operations, according to jsPerf tests.
It also seems to be supported on IE6.
function replaceContentInContainer(matchClass, content) {
var re = new RegExp("(?:^|\\s)" + matchClass + "(?!\\S)"),
elems = document.getElementsByTagName('*'), i;
for (i in elems) {
if (re.test(elems[i].className)) {
elems[i].innerHTML = content;
}
}
}
replaceContentInContainer("box", "This is the replacement text.");
If you look for the same class(es) frequently, you can further improve it by storing the (precompiled) regular expressions elsewhere, and passing them directly to the function, instead of a string.
function replaceContentInContainer(reClass, content) {
var elems = document.getElementsByTagName('*'), i;
for (i in elems) {
if (reClass.test(elems[i].className)) {
elems[i].innerHTML = content;
}
}
}
var reBox = /(?:^|\s)box(?!\S)/;
replaceContentInContainer(reBox, "This is the replacement text.");

This should work in pretty much any browser...
function getByClass (className, parent) {
parent || (parent=document);
var descendants=parent.getElementsByTagName('*'), i=-1, e, result=[];
while (e=descendants[++i]) {
((' '+(e['class']||e.className)+' ').indexOf(' '+className+' ') > -1) && result.push(e);
}
return result;
}
You should be able to use it like this:
function replaceInClass (className, content) {
var nodes = getByClass(className), i=-1, node;
while (node=nodes[++i]) node.innerHTML = content;
}

var elems = document.querySelectorAll('.one');
for (var i = 0; i < elems.length; i++) {
elems[i].innerHTML = 'content';
};

I assume this was not a valid option when this was originally asked, but you can now use document.getElementsByClassName('');. For example:
var elements = document.getElementsByClassName(names); // or:
var elements = rootElement.getElementsByClassName(names);
See the MDN documentation for more.

There are 3 different ways to get elements by class in javascript. But here for your query as you have multiple elements with the same class names you can use 2 methods:
getElementsByClassName Method - It returns all the elements with the specified class present in the document or within the parent element which called it.
function ReplaceContentInContainer(className, content) {
var containers = document.getElementsByClassName(className);
for (let i = 0; i < containers.length; i++) {
containers[i].innerHTML = content;
}
}
ReplaceContentInContainer('box', 'This is the replacement text');
<div class='box'></div>
querySelectorAll Method - It select element on the basic of CSS selectors. Pass your CSS class to it with a dot and it will return all the element having specified class as an array-like object.
function ReplaceContentInContainer(className, content) {
var containers = document.querySelectorAll(`.${className}`);
for (let i = 0; i < containers.length; i++) {
containers[i].innerHTML = content;
}
}
ReplaceContentInContainer('box', 'This is the replacement text');
<div class='box'></div>

I think something like:
function ReplaceContentInContainer(klass,content) {
var elems = document.getElementsByTagName('*');
for (i in elems){
if(elems[i].getAttribute('class') == klass || elems[i].getAttribute('className') == klass){
elems[i].innerHTML = content;
}
}
}
would work

jQuery handles this easy.
let element = $(.myclass);
element.html("Some string");
It changes all the .myclass elements to that text.

When some elements lack ID, I use jQuery like this:
$(document).ready(function()
{
$('.myclass').attr('id', 'myid');
});
This might be a strange solution, but maybe someone find it useful.

Select Element By CSS style (all with given style)

Is there a way to select all elements that have a given style using JavaScript?
Eg, I want all absolutely positioned elements on a page.
I would assume it is easier to find elements by style where the style is explicitly declared:
the style is non-inherited (such as positioning)
the style is not the default (as would be position:static).
Am I limited to those rules? Is there a better method for when those rules apply?
I would happily to use a selector engine if this is provided by one (ideally Slick - Mootools 1.3)
EDIT:
I came up with a solution that will only work with above rules.
It works by cycling through every style rule, and then selector on page.
Could anyone tell me if this is better that cycling through all elements (as recommended in all solutions).
I am aware that in IE I must change the style to lowercase, but that I could parse all styles at once using cssText. Left that out for simplicity.
Looking for best practice.
var classes = '';
Array.each(documents.stylesheets, function(sheet){
Array.each(sheet.rules || sheet.cssRules, function(rule){
if (rule.style.position == 'fixed') classes += rule.selectorText + ',';
});
});
var styleEls = $$(classes).combine($$('[style*=fixed]'));

You can keep Mootools, or whatever you use... :)
function getStyle(el, prop) {
var view = document.defaultView;
if (view && view.getComputedStyle) {
return view.getComputedStyle(el, null)[prop];
}
return el.currentStyle[prop];
}
​function getElementByStyle(style, value, tag)​ {
var all = document.getElementsByTagName(tag || "*");
var len = all.length;
var result = [];
for ( var i = 0; i < len; i++ ) {
if ( getStyle(all[i], style) === value )
result.push(all[i]);
}
return result;
}

For Mootools:
var styleEls = $$('*').filter(function(item) {
return item.getStyle('position') == 'absolute';
});

In jQuery you could use
$('*').filter( function(){
return ($(this).css('position') == 'absolute');
} );
[update]
Or even create a new selector.
got me interested and so here is one (its my 1st, so its not built for efficiency) to find elements by css property..
$.expr[':'].css = function(obj, index, meta, stack){
var params = meta[3].split(',');
return ($(obj).css(params[0]) == params[1]);
};
usage: $('optionalSelector:css(property,value)')
will return all elements (of optionalSelector) whose property = value
example: var visibleDivs = $('div:css(visibility,visible)');
will return all divs whose visibility is set to visible (works for the default visibility as well..)

There is no selector for CSS attributes, so you're pretty much stuck to looping through each element and checking it's position. Here's a jQuery method:
$("*").each(function() {
var pos = $(this).css('position');
if(pos == "absolute") {
// do something
}
else if (pos == "relative") {
// do something else
}
});
You can use Case statements instead of if/else as well.
Other than this solution, there is no selector per se that can search by CSS attributes (unless they were inline, maybe).