Good points about the zip code, and running $.query() all the time, Mike.
Maybe the two approaches can be combined? Run the code to munge the
query string the first time $.query() is called. Cache two sets of
results -- a 'raw' set and a 'type cast' set -- in a closure, and return
one of the cached objects on subsequent calls to $.query()
(function(){
var after_type_cast = {};
var before_type_cast = {};
var cached = false;
jQuery.query = function(cast) {
if(!cached) {
// remove leading ? and trailing &
var q =
location.search.replace(/^\?/,'').replace(/\&$/,'').split('&');
for( var i = q.length - 1; i >= 0; i-- ) {
var p = q[i].split('='), key = p[0], val = p[1];
before_type_cast[key] = val;
// convert floats
if(/^[0-9.]+$/.test(val))
val = parseFloat(val);
// convert booleans
if(/^(true|false)$/.test(val))
val = (val == 'true');
// ingnore empty values
if(val)
after_type_cast[key] = val;
}
cached = true;
}
return cast === false ? before_type_cast : after_type_cast;
};
})();
I haven't tested the above, and it can probably be condensed/simplified
a bit, but you get the idea ...
So, for http://www.example.com/test?special=&zip=00125, pass false to
$.query() to get the raw values ...
$.query().zip -> 125 (number)
$.query(false).zip -> 00125 (string)
$.query().hasOwnProperty('special') -> false
$.query(false).hasOwnProperty('special') -> true
Luke
Michael Geary wrote:
> In Jörn's version of the code, you wouldn't call $.query() at all. Instead,
> $.query would be the resulting object that you would then use directly. Note
> that jQuery and $ are references to the same object, so jQuery.query and
> $.query are the same thing.
>
> I'm not sure I like this idea since it means the query parsing would be done
> whether you need it or not. But maybe it doesn't take enough time to worry
> about. If it does run all the time, it would be worth optimizing the code by
> using a conventional for loop instead of jQuery.each and combining the
> repeated "split" calls.
>
> I'd also remove the code that does the float conversion and ignores empty
> values. Consider this URL:
>
> http://www.example.com/test?special=&zip=00125
>
> I may want to know about that "special" parameter, and I probably don't want
> the Zip code turned into the number 125.
>
> Here's how I would code it...
>
> For the "jQuery.query is an object" version:
>
> (function() {
> var r = jQuery.query = {};
> var params = location.search.replace(/^\?/,'').split('&');
> for( var i = params.length-1; i >= 0; i-- ) {
> var p = params[i].split('='), key = p[0];
> if( key ) r[key] = p[1];
> }
> })();
>
> For the "jQuery.query is a function":
>
> jQuery.query = function() {
> var r = {};
> var params = location.search.replace(/^\?/,'').split('&');
> for( var i = params.length-1; i >= 0; i-- ) {
> var p = params[i].split('='), key = p[0];
> if( key ) r[key] = p[1];
> }
> return r;
> };
>
> All untested, of course. :-)
>
> -Mike
_______________________________________________
jQuery mailing list
[email protected]
http://jquery.com/discuss/
