Contextual Regular Expression - regex

Question

I have a list of comma separated words that I want to remove the comma from and replace with a space:
elements-(a,b,c,d)
becomes:
elements-(a b c d)
The question is how can I do this using a regular expression if and only if that list is within a specific context, e.g. only prefixed by element-():
The following:
There are a number of elements-(a,b,c,d) and a number of other elements-(e,f,g,h)
should become:
There are a number of elements-(a b c d) and a number of other elements-(e f g h)
What would be the correct way to do this with regex?

Answer 1

For contextual regular expressions, you can use zero-width look-around assertions. Look-around assertions are used to assert that something must be true in order for the match to succeed, but they do not consume any characters (hence "zero-width").
In your case, you want to use positive look-behind and look-ahead assertions. In C#, you can do the following:
static string Replace(string text)
{
return Regex.Replace(
text,
#"(?<=elements\-\((\w+,)*)(\w+),(?=(\w+,)*\w+\))",
"$2 "
);
}
There are three basic parts to the pattern here (in order):
(?<=elements\-\((\w+,)*) - this is the positive look-behind assertion. It says that the pattern will only match if it is preceded by the text elements-( and zero-or-more comma-separated strings.
(\w+), - this is the actual match. It's the text that's being replaced.
(?=(\w+,)*\w+\)) - this is the positive look-ahead assertion. It says that the pattern will only match if it is followed by one-or-more comma-separated strings.
In C#, for matching the inner comma-separated contents, you can alternatively do the following:
static string Replace(string text)
{
return Regex.Replace(
text,
#"(?<=elements\-)\(((\w+,)+\w+)\)",
m => string.Format("({0})", m.Groups[1].Value.Replace(',', ' '))
);
}
The basic approach with the positive look-ahead assertion is still the same.
Example output:
"(x,y,z) elements-(a,b) (m,m,m) elements-(c,d,e,f,g,h)"
...becomes...
"(x,y,z) elements-(a b) (m,m,m) elements-(c d e f g h)"