Skip to content

Instantly share code, notes, and snippets.

@lsmith
Created June 15, 2011 17:51
Show Gist options
  • Save lsmith/1027653 to your computer and use it in GitHub Desktop.
Save lsmith/1027653 to your computer and use it in GitHub Desktop.
Y.config.base = YUI.config.base || Y.Env.getBase(Y.Env._BASE_RE);
// Regex in English:
// I'll start at the \b(simpleyui).
// 1. Look in the test string for "simpleyui" or "yui" or
// "yui-base" or "yui-rls" or "yui-foobar" that comes after a word break. That is, it
// can't match "foyui" or "i_heart_simpleyui". This can be anywhere in the string.
// 2. After #1 must come a forward slash followed by the string matched in #1, so
// "yui-base/yui-base" or "simpleyui/simpleyui" or "yui-pants/yui-pants".
// 3. The second occurence of the #1 token can optionally be followed by "-debug" or "-min",
// so "yui/yui-min", "yui/yui-debug", "yui-base/yui-base-debug". NOT "yui/yui-tshirt".
// 4. This is followed by ".js", so "yui/yui.js", "simpleyui/simpleyui-min.js"
// 0. Going back to the beginning, now. If all that stuff in 1-4 comes after a "?" in the string,
// then capture the junk between the LAST "&" and the string in 1-4. So
// "blah?foo/yui/yui.js" will capture "foo/" and "blah?some/thing.js&3.3.0/build/yui-rls/yui-rls.js"
// will capture "3.3.0/build/"
//
// Regex Exploded:
// (?:\? Find a ?
// (?:[^&]*&) followed by 0..n characters followed by an &
// * in fact, find as many sets of characters followed by a & as you can
// ([^&]*) capture the stuff after the last & in \1
// )? but it's ok if all this ?junk&more_junk stuff isn't even there
// \b(simpleyui| after a word break find either the string "simpleyui" or
// yui(?:-\w+)? the string "yui" optionally followed by a -, then more characters
// ) and store the simpleyui or yui-* string in \2
// \/\2 then comes a / followed by the simpleyui or yui-* string in \2
// (?:-(min|debug))? optionally followed by "-min" or "-debug"
// .js and ending in ".js"
Y.Env._BASE_RE = /(?:\?(?:[^&]*&)*([^&]*))?\b(simpleyui|yui(?:-\w+)?)\/\2(?:-(min|debug))?\.js/;
...
getBase: G_ENV && G_ENV.getBase ||
function(pattern) {
var nodes (doc && doc.getElementsByTagName('script')) || [],
path = Env.cdn,
i, len, src, match;
for (i = 0, len = nodes.length; i < len; ++i) {
src = nodes[i].src;
if (src) {
match = src.match(pattern);
if (match) {
path = RegExp.leftContext || src.slice(0, src.indexOf(match[0]));
// this is to set up the path to the loader. The file
// filter for loader should match the yui include.
filter = match[3];
// extract correct path for mixed combo urls
// http://yuilibrary.com/projects/yui3/ticket/2528423
if (match[1]) {
path += '?' + match[1];
}
break;
}
}
}
// use CDN default
return path;
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment