normalizeDateSeparators function

String normalizeDateSeparators(String line)

Removes OCR-introduced spaces around date separators and digit clusters.

Implementation

String normalizeDateSeparators(String line) {
  if (line.isEmpty) {
    return line;
  }

  String value = line.replaceAllMapped(
    RegExp(r'(?<=\d)\s*([./-])\s*(?=\d)'),
    (Match match) => match.group(1) ?? '',
  );
  value = value.replaceAllMapped(RegExp(r'(?<=\d)\s*,\s*(?=\d)'), (_) => ',');

  // Collapse spaces around dots between digits and alphanumeric characters.
  value = value.replaceAllMapped(
    RegExp(r'(?<=\d)\s*\.\s*(?=[A-Za-z0-9])'),
    (_) => '.',
  );
  value = value.replaceAllMapped(
    RegExp(r'(?<=[A-Za-z0-9])\s*\.\s*(?=\d)'),
    (_) => '.',
  );

  // Collapse split numeric runs like "1 2 3" only when there are 2+ joins
  // and all digit groups are single digits (fragmented number, not a column
  // of distinct numbers).
  final RegExp splitDigits = RegExp(r'(?<=\d)\s+(?=\d)');
  final int joins = splitDigits.allMatches(value).length;
  if (joins >= _digitJoinMinCount && _allSingleDigitGroups(value)) {
    value = value.replaceAll(splitDigits, '');
  }
  return value;
}