正则表达式将文本与方括号匹配

Question

Using javascript I need to get the text and tokens from string like:

"type":"homePhone","means":"$[createJohnRequest.contactInfo[type=homePhone].means]"

Such that a regex will return:

$[createJohnRequest.contactInfo[type=homePhone].means]

I have a few attempts at this but none that work:

/(\$\[(.*?]))/g  

will return: $[createJohnRequest.contactInfo[type=homePhone]

/(\$\[(.*]))/g 

This works in the above case but is way too greedy for a case like:

{"firstName":"$[user.firstName]","userName":"$[user.username1]","details":
{"description":"this is $[user.username1] the $[user.username2] text th $[user.username3] 
at conta$[user.username4]ins the terms we want to find. $[final.object]"}}

Ideally I want a single regex to match both in multiline text:

some text here $[some.value.here]bunch of noisy text in between here
some more text here$[some.value[index]goes.here]some more noise here

$[some.value.here] and $[some.value[index].goes.here]

Anyone have any ideas to point me in the right direction?

I am leaning toward using $[some token]$ instead which is pretty simple to capture.

Answer 1

You want something like this, as long as the bracket nesting level is limited to two:

/\$\[(\[.*]|.)*?\]/g

In English:

a dollar sign followed by 
a bracketed sequence whose content is
    any number of occurrences of 
        either 
            a bracketed subsequence whose content is anything
            or something else

See http://regexr.com/39iht .

If you want to support deeper nesting, I suggest you write a little function to build the regexp for you, so you don't go insane:

function make_regexp(level) {
    var build = "\\[.*?]";
    while (level--) {
        build = "(\\[" + build + "]|.)*?";
    }
    return RegExp("\\$" + build, "g");
}

> make_regexp(3)
/\$(\[(\[(\[.*]|.*)*?]|.*)*?]|.*)*?/g