正则表达式从一个通用正则表达式中获取多个捕获组

Question

This is for JavaScript.

My Problem:

So I want a regex that can get multiple sub capture groups for a string. Right now my attempt only gets one, while I expect 3. Ideally, it should be varied in that it should be able to get 2 or 3 or more.

My attempt

 let characters = { some: { test: { var: "hi" } } } let b = document.getElementById("body"); let h = b.innerHTML; let regex = /(?:\\$\\{)([^.]*[.]?)*(?:\\})/; let match; match = regex.exec(h); console.log(match); 

 <div id="body"> ${some.test.var} </div> 

What I want

 let characters = { some: { test: { var: "hi" } } } let b = document.getElementById("body"); let h = b.innerHTML; let regex = /(?:\\$\\{)([^.]*)[.]([^.]*)[.]([^.]*)(?:\\})/; let match; match = regex.exec(h); console.log(match); 

 <div id="body"> ${some.test.var} </div> 

Answer 1

It seems you want to use a single capturing group to obtain multiple groups in the resulting array. It is not possible with JavaScript regex. See Repeating a Capturing Group vs. Capturing a Repeated Group . There will always be as many groups in the resulting array as there are capturing groups in the pattern (+1, the zeroth group containing the whole match). If you use (\\w* )* it will match 0+ word chars and then a space and place it into Group 1 buffer, then, if it finds another streak of word chars and a space, the regex engine will match this text portion, replace the contents inside Group 1, and will go on searching, etc.

You need to match ${...} substrings first (there may be several options here, BTW, I suggest /\\${(\\w+(?:\\.\\w+)*)}/g to only match ${...} that have word chars and . inside, but it can also be /\\${([^}]+)}/g ) and then split with . :

 const regex = /\\${(\\w+(?:\\.\\w+)*)}/g; const str = `\\${some.test.var} \\${some.test.var.more.here}`; let m; while (m = regex.exec(str)) { console.log(m[1].split('.')); } 

Note you have the whole match value in m[0] .

Pattern details :

• \\${ - a ${ literal substring
• (\\w+(?:\\.\\w+)*) - Group 1 capturing:
  • \\w+ - 1 or more word chars
  • (?:\\.\\w+)* - zero or more sequences of:
    • \\. - a dot
    • \\w+ - 1 or more word chars
• } - a literal }

If you want to simplify the pattern, use /\\${([.\\w]+)}/g or /\\${([^}]+)}/g .