What does \b really mean in Ruby regular expressions?

Question

I have a file with phrases such as "Canyon St / 27th Way" that I am trying to turn into "Canyon St and 27th Way" with Ruby regular expressions.

I used file = file.gsub(/(\b) \/ (\b)/, "#{$1} and #{$2}") to make the match, but I am a little stumped about what \b really means and why $1 contains all of the characters before the word boundary that precedes the slash and why $2 contains all of the characters after the word boundary starting the next word.

Usually, I expect that whatever is in parentheses in a regular expression would be in $1 and $2, but I am not sure what parentheses around a word boundary would really mean because there really is nothing between the transition from a word character to a white space character.

Answer 1

The parentheses aren't doing anything in this context. You could get the same result using /\b \/ \b/.

I think you are getting a little confused by $1 and $2. Those aren't actually doing anything either. They are nil because they are matching nothing (just a word boundry). What you have written is the logical equivalent of .gsub(/\b \/ \b/, " and ")

Answer 2

The $1 and $2 are not actually related to your regex match: a method's arguments are evaluated before the method is called, so

"#{$1} and #{$2}"

Is evaluated before the regex is matched against your string. If you haven't done earlier regex matches then these variables will be nil, so you're actually doing

file = file.gsub(/(\b) \/ (\b)/, " and ")

that is you are replacing a slash surrounded by spaces by "and", also surrounded by spaces. $1 and $2 will be updated to be empty strings, and so you'll see the same behaviour when you process the next string.