I have the following strings:

| Elencia   { 1: Simple one, 0: A longer sentence, or maybe not, 2: Another sentence }
| Elencia   { 10:    Sentence with spaces on left,     0: Same but with some spaces on right      }

I would like to catch the name on the left and the number the sentence avoiding the multiple spaces on the left and the right.

I wrote this at the moment:

(?<point>[\d] ):\s (?<content>[\w ] )

but it only catches the part inside the curly brackets and it includes the spaces at the end, which is not what I want.

What I would like to have. First example:

- Elencia as name
- "1" as point, "Simple one" as sentence
- "0" as point, "A longer sentence, or maybe not" as sentence
- "2" as point, "Another sentence" as sentence

Second example:

- Elencia as name
- "10" as point, "Sentence with spaces on left" as sentence
- "0" as point, "Sentence with spaces on left" as sentence

I'm using the regex with gdscript but PHP or JavaScript style would work too.

Thank you for your help!

CodePudding user response：

You may use this regex:

(?:(?<name>\w )\s {|(?!^)\G)\W*(?<point>\d ):\s (?<content>. ?(?=,\s \d :|\s*}))

RegEx Demo

RegEx Details:

• (?:: Start a non-capture group
  • (?<name>\w ): Match 1 word characters in group named name
  • \s {: Match 1 whitespace followed by a {
  • |: OR
  • (?!^)\G: Start from end of the previous match
• ): End non-capture group
• \W*: Match 0 non-word characters
• (?<point>\d ): Match 1 digits in names capture group point
• :\s : Match a : followed by 1 whitespace
• (?<content>. ?(?=,\s \d:|\s*}):
• (?=: Start a positive lookahead assertion
  • ,\s \d :: Match comma, 1 whitespace, 1 digits followed by :
  • |: OR
  • \s*}: Match 0 whitespace followed by }
• ): End positive lookahead assertion