PlatformTargetAnalyzer.fromContent constructor

PlatformTargetAnalyzer.fromContent(String content)

Parses the annotation from already-loaded content (zero file reads).

Uses balanced-paren scanning (not a [^)]+ regex) to find the matching close paren — a plain [^)]+ stops at the FIRST ) regardless of nesting, silently truncating the captured annotation (and every getter's regex along with it) the moment any ) appears before the real end — including inside an ordinary parenthesized comment like // (see the docs) sitting between two annotation params. Found via a real, self-inflicted repro: adding exactly such a comment to nitro_type_coverage's @NitroModule made every getter on this class silently return false.

Implementation

factory PlatformTargetAnalyzer.fromContent(String content) {
  final startMatch = RegExp(r'@NitroModule\s*\(').firstMatch(content);
  if (startMatch == null) return PlatformTargetAnalyzer._('');
  final body = _scanBalancedParens(content, startMatch.end);
  return PlatformTargetAnalyzer._(body.replaceAll('\n', ' '));
}