How to Generate All Possible CSS 2 Selector Combinations?_问答_开发者

How to Generate All Possible CSS 2 Selector Combinations?

开发者 https://www.devze.com 2023-02-14 05:53 出处：网络

What would be the best way to generate all possible CSS 2 Selector Combinations for a DOM element in context with the current state of the document?

What would be the best way to generate all possible CSS 2 Selector Combinations for a DOM element in context with the current state of the document?

Eg: For the following MarkUp

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en">
<head>
  <meta http-equiv="Content-Type" content="text/html;charset=UTF-8" />
  <title></title>
</head>
<body>
  <div id="content">
    <ul>
      <li>a</li>
      <li>b</li>
      &l开发者_StackOverflowt;li class="last-li">c</li> <!--ARGUMENT ELEMENT-->
    </ul>
  </div>
</body>
</html>

A JS/jQuery Method such that the  in the above markup is given as an argument to the method it returns all possible CSS 2 Selector combinations for the Argument in form of an array. An optional parameter can be added to the method which will define the maximum depth of CSS 2 Selector Nesting. If this parameter is set to false it should return all possible combinations (Will be a performance hogger if set to false)

Example Output:

[
 ".last-li",
 "li.last-li",
 "ul .last-li",
 "ul li.last-li",
 "div ul .last-li",
 "div ul li.last-li",
 "#content ul .last-li",
 "#content ul li.last-li",
 "div#content ul .last-li",
 "div#content ul li.last-li",
 "body div ul li.last-li",
 "body div ul .last-li",
 "body #content ul li.last-li",
 "body div#content ul .last-li",
 "body div#content ul li.last-li",
]

Any pointers in this regard will be extremely helpful.

First, let's stick to a narrow class of selectors involving tag names, class names and IDs, nothing fancy like E > F or E + F. Let's also disallow combinations of class names (.class1.class2.class3), otherwise a single element with 10 class names would generate 4 million selectors alone.

Each of our full selectors consists of simple selectors separated by spaces. Every simple selector is a combination of tag{0,1}id{0,1}class{0,n} - i.e. each element has exactly one tag, at most one ID, and it can have an arbitrary number of class names. That gives us the upper limit of 2 * 2 * (n + 1) simple selectors for a single element.

Given a reference to a DOM element grab it's tag name, ID and class names. Calculate all possible simple selectors as described above. Lets call this set A1. Move one step up the hierarchy to it's parent, calculate all simple selectors for that parent element - that'll be the set A2. Continue until you reach the html element - the set Am. Now you'll have a list, consisting of m items, each item is a set of simple selectors.

Now pick some of these sets and find their cartesian product. Say, m = 5. How many sets can you pick? The set A1 is always present, but others are optional. For each one of them you either pick it or not. That's like binary numbers:

0000 // 0, A1
0001 // 1, A1 x A2
0010 // 2, A1 x A3
0011 // 3, A1 x A2 x A3
0100 // 4, A1 x A4
...

That means you'll have 2^(m-1) cartesian products. You can now convert them to strings. The last step is to remove duplicates, consider this example:

<div>
  <div>
    <span></span>
  </div>
</div>

Our calculations will yield this list:

  span
  div span // inner div
  div span // outer div
  div div span

Those two divs yield duplicate selectors. Remove those and the job is done. All the steps are very simple algorithmically. I'm sure you can figure them out, but if you get stuck somewhere or need further clarification feel free to ask me in the comments.

UPDATE

So, I decided to play with it a bit more and wrote the program, here's the list of selectors your example generates: http://pastie.org/1616164