Rewrite regex for newer rubys

ruby -wc now shows context lines and a prettier format for syntax
errors.  Adjust the regex for that.

Note that ruby might emit "syntax errors" (plural!) for one line, and
we just read and pass the first one.

Author: kaste

linter.py

@@ -1,9 +1,9 @@
 #
 # linter.py
-# Linter for SublimeLinter3, a code checking framework for Sublime Text 3
+# Linter for SublimeLinter, a code checking framework for Sublime Text
 #
 # Written by Aparajita Fishman
-# Copyright (c) 2015-2016 The SublimeLinter Community
+# Copyright (c) 2015-2025 The SublimeLinter Community
 # Copyright (c) 2013-2014 Aparajita Fishman
 #
 # License: MIT
@@ -23,56 +23,63 @@ class Ruby(RubyLinter):
     }
 
     cmd = 'ruby -wc'
-    regex = (
-        r'^(?P<file>.+?):(?P<line>\d+): (?:(?P<error>.*?error)|(?P<warning>warning))[,:] (?P<message>[^\r\n]+)\r?\n'
-        r'(?:^[^\r\n]+\r?\n^(?P<col>.*?)\^)?'
-    )
-    multiline = True
-    comment_re = r'\s*#'
+    regex = r"""
+        (?xm)
+        ^(?:(?P<exe>.*:\s)?)?                # optional "C:/ruby...:" prefix
+        (?P<filename>.+?):(?P<line>\d+):     # file and line
+        (?:\s?(?:(?P<error>.*?error)s?|(?P<warning>warning))) # "syntax error" or warning
+        [,:]?\s(?P<message>[^\n]+)           # headline message
+        (?:\n                                # optional multiline block
+            (?:
+              (?:\s*\d+\s*\|[^\n]*\n)*       # optional leading context lines
+              >\s*(?P<culprit_line>\d+)\s*\|[^\n]*\n # culprit line
+              \s*\|\s(?P<col>.*?)\^\s        # caret line with message
+              (?P<culprit_message>[^\n]*)    # rest of caret line (error text)
+            )?
+        )?
+    """
     on_stderr = None
 
     def split_match(self, match):
         """
-        Return the components of the match.
-
-        We override this because unrelated library files can throw errors,
-        and we only want errors from the linted file.
+        Return the error as matched by the regex.
 
+        We override this for a better error message.
         """
 
-        if match:
-            if match.group('file') != '-':
-                match = None
-
-        match, line, col, error, warning, message, _ = super().split_match(match)
-        near = self.search_token(message)
+        error = super().split_match(match)
+        if culprit_message := error.get("culprit_message", None):
+            error["message"] = culprit_message
+        error["near"] = self.search_token(error.message)
+        if error.error:
+            error["error"] = error.error.replace(" ", "-")
 
-        return match, line, col, error, warning, message, near
+        return error
 
     def search_token(self, message):
         """Search text token to be highlighted."""
 
         # First search for variable name enclosed in quotes
-        m = re.search("(?<=`).*(?=')", message)
+        m = re.search(r"(?<=`).*(?=')", message)
 
         # Then search for variable name following a dash
         if m is None:
-            m = re.search('(?<= - )\S+', message)
+            m = re.search(r'(?<= - )\S+', message)
 
         # Then search for mismatched indentation
         if m is None:
-            m = re.search("(?<=mismatched indentations at ')end", message)
+            m = re.search(r"(?<=mismatched indentations at ')end", message)
 
         # Then search for equal operator in conditional
         if m is None:
-            m = re.search('(?<=found )=(?= in conditional)', message)
+            m = re.search(r'(?<=found )=(?= in conditional)', message)
 
         # Then search for use of operator in void context
         if m is None:
-            m = re.search('\S+(?= in void context)', message)
+            m = re.search(r'\S+(?= in void context)', message)
 
         # Then search for END in method
         if m is None:
-            m = re.search('END(?= in method)', message)
+            m = re.search(r'END(?= in method)', message)
 
         return m.group(0) if m else None