JavaScript a question on string decomposition_javascript skills

I went to a certain company for an interview (I won’t mention the company name, but this set of questions may still be used). I did a set of questions in 30 minutes on site, and one of them was like this:

Required to write a function in js , for the incoming URL string in the following form, return the corresponding object.
For example:
If the string a='?name=zhiyelee&blog=www.tsnrose.com';
is returned, b={'name':'zhiyelee','blog':'www .tsnrose.com'}


Due to the relatively short time at that time, there were some problems with the implementation. After I came back, I thought about it and summarized it as follows:
I thought of two ideas, one is to use regular expressions Expression, the second is to use the split function of string.

1. Use regular expression processing
The first thing I thought of was to use regular expression processing. It may be that this is more challenging and the simplest to write. However, I doubted that the efficiency of this method would be It is less efficient than directly using string functions. We will verify this efficiency below~
The idea of ​​this method is very simple, which is to use regular expressions to match a string of '***=###' each time, and then loop and finally remove them all.
The code is as follows

Copy code The code is as follows:

var getNRReg = function(str) {
var res = {};
var reg = /(w )=(w )/g;
while ((a = reg.exec(str))) {
res[a [1]] = a[2];
}
return res;
};

2. The idea of ​​using string functions to process
is also relatively simple Conventional, but it is definitely more troublesome to write than using regular expressions.
My idea is to first use '&' to split the original string into multiple strings. The style of each string is like '***=###', and then apply split('=') to the string. . (This method does not use regular expressions, of course we can use regular split(/&|=/)
The code is as follows

Copy code The code is as follows:

var getNRSplit = function(str) {
var temp, res = {},
i, ret;
str = str.slice(1 );
temp = str.split('&');
for (i = 0; i ret = temp[i].split('=' );
res[ret[0]] = ret[1];
}
return res;
}

Efficiency issues of these two methods
It is well known that the efficiency of regular expressions is a bit low, so I initially thought that the efficiency of the first solution must be lower than the efficiency of the second solution, so I measured it and found otherwise:

Time-consuming to execute 1,000,000 times Situation:
getNRReg execution takes 4399ms
getNRSplit execution takes 6116ms

View the complete source code: jsfiddle, you can test it by yourself~

ps:
Finally attached is a written test question from Sina Weibo 2011-06-15 for a front-end position at Beijing Post Information Conference:

Use regular expressions to delete characters that are adjacent and identical to a certain character. For example, the string "fdaffdaaklfjk" becomes "fdafdakljk" after processing.

For my answer see jsfiddle