正则表达式对字符串中的递归模式进行分组并将捕获组作为数组返回

Question

I have a string that looks like this: Hello:A1,A3,A4,AA04

Right now I have a regex query that looks like this:

/(?<g1>[A-Za-z]+):(?<g2>[A-Za-z0-9,]+)/

When a match is found I can access it like this:

"Hello:A1,A3,A4,AA04".match(/(?<g1>[A-Za-z]+):(?<g2>[A-Za-z0-9,]+)/).groups.g1 // outputs: Hello

ok...now I also have g2 . I need g2 in an array with the , gone. Right now in my code I have done it this way:

"Hello:A1,A3,A4,AA04".match(/(?<g1>[A-Za-z]+):(?<g2>[A-Za-z0-9,]+)/).groups.g2.split(',') // outputs: [ 'A1', 'A3', 'A4', 'AA04' ]

Now the question is: Is it possible to accomplish this without .split(',') by using Regex and have it return like this:

"Hello:A1,A3,A4,AA04".match(?).groups.g2 // would output: [ 'A1', 'A3', 'A4', 'AA04' ]

I know I'm not the best at explaining this but I tried to find an answer, I'm either missing it or it's not possible. This is not critical or super important just something on my mind. Sorry for the horrible title.

running Node.js v13.2.0

Answer 1

Added a ^ to match beginning of string only. Using lookaheads.
Note that lookaheads doesn't have good browser support.
Also, I don't know/haven't tested the performance characteristics of this versus exec looping. And to be entirely honest, it's not much cleaner looking than split. If you want a one-liner, this'll suffice I guess.
I remember for very long strings on other platforms using RegExs like this didn't do well. But I ended up resorting to other methods to searching anyways.

 console.log( "Hello:A1,A3,A4,AA04".match(/(^[A-Za-z]+(?=:))|((?=,?)([A-Za-z0-9]+))/g) )