I would like to remove an unknown substring when it occurs between two known substrings (<foo> and </foo>). For example, I'd like to convert:
hello <foo>remove me</foo>
to:
hello <foo></foo>
I can do it with:
s = ...
s.replace(/<foo>.*?<\/foo>/, '<foo></foo>')
but I'd like to know if there's a way to do it without repeating the known substrings (<foo> and </foo>) in the regex and the replacement text.
You can capture tag in a captured group and use it later as back reference:
var repl = s.replace(/<(foo)>.*?<\/\1>/, '<$1></$1>');
//=> hello <foo></foo>
Note \1 and $1 are back references to the captured group #1.
Try below regex using grouping.
(?:<foo>)(.*?<\/foo>)
Pictorial representation: Debuggex Demo
Sample code:
var re = /(?:<foo>)(.*?<\/foo>)/;
var str = 'hello <foo>remove me</foo>';
var subst = '<foo></foo>';
var result = str.replace(re, subst);
Output:
hello <foo></foo>
"<foo>abcdefg</foo><foo>hijklmno</foo>" since it will match the entire string instead of <foo>abcdefg</foo> because the quantifier * is greedy. About, regex101 steps (or the regexbuddy steps): It is usefull to see how many steps are necessary to make your pattern succeed or fail. But keep in mind that this information is not always related to the real pattern performance. The only way to be sure is a timer.