encodeHeader method

@override String encodeHeader(String text, { int nameLength = 0, Codec codec = utf8, bool fromStart = false, })

Encodes the header text in Q encoding only if required. Compare https://tools.ietf.org/html/rfc2047#section-4.2 for details. text specifies the text to be encoded. nameLength the length of the header name, for calculating the wrapping point. codec the optional codec, which defaults to utf8. Set the optional fromStart to true in case the encoding should start at the beginning of the text and not in the middle.

Implementation

@override
String encodeHeader(final String text,
    {int nameLength = 0, Codec codec = utf8, bool fromStart = false}) {
  var runes = List.from(text.runes, growable: false);
  var numberOfRunesAbove7Bit = 0;
  var startIndex = -1;
  var endIndex = -1;
  final runeCount = runes.length;

  for (var runeIndex = 0; runeIndex < runeCount; runeIndex++) {
    var rune = runes[runeIndex];
    if (rune > 128) {
      numberOfRunesAbove7Bit++;
      if (startIndex == -1) {
        startIndex = runeIndex;
        endIndex = runeIndex;
      } else {
        endIndex = runeIndex;
      }
    }
  }
  if (numberOfRunesAbove7Bit == 0) {
    return text;
  } else {
    // TODO Set the correct encoding
    final qpWordHead = '=?utf8?Q?';
    final qpWordTail = '?=';
    final qpWordDelimSize = qpWordHead.length + qpWordTail.length;
    if (fromStart) {
      startIndex = 0;
      endIndex = text.length - 1;
    }
    // Available space for the current encoded word
    var qpWordSize = MailConventions.encodedWordMaxLength -
        qpWordDelimSize -
        startIndex -
        (nameLength + 2);
    // Counts the characters of the current encoded word
    var wordCounter = 0;
    // True when reached the end of the current word available space
    var isWordSplit = false;
    var buffer = StringBuffer();
    for (var runeIndex = 0; runeIndex < runeCount; runeIndex++) {
      var rune = runes[runeIndex];
      if (runeIndex < startIndex || runeIndex > endIndex) {
        buffer.writeCharCode(rune);
        continue;
      }
      if (runeIndex == startIndex || isWordSplit) {
        // Adds the line terminator
        if (isWordSplit) {
          buffer
            ..write(qpWordTail)
            // NOTE Per specification, a CRLF should be inserted here,
            // but the folding occurs on the rendering function.
            // Here we leave only the WSP marker to separate each q-encode word.
            // ..writeCharCode(AsciiRunes.runeCarriageReturn)
            // ..writeCharCode(AsciiRunes.runeLineFeed)
            // Assumes per default a single leading space for header folding
            ..writeCharCode(AsciiRunes.runeSpace);
          // Resets the split flag
          isWordSplit = false;
          // Calculates the new encoded word size
          qpWordSize =
              MailConventions.encodedWordMaxLength - qpWordDelimSize - 1;
        }
        buffer.write(qpWordHead);
      }
      if ((rune > AsciiRunes.runeSpace && rune <= 60) ||
          (rune == 62) ||
          (rune > 63 && rune <= 126 && rune != AsciiRunes.runeUnderline)) {
        wordCounter++;
        isWordSplit = wordCounter > qpWordSize;
        if (!isWordSplit) {
          buffer.writeCharCode(rune);
        }
      } else if (rune == AsciiRunes.runeSpace) {
        wordCounter++;
        isWordSplit = wordCounter > qpWordSize;
        if (!isWordSplit) {
          buffer.write('_');
        }
      } else {
        // _writeQuotedPrintable(rune, buffer, codec);
        final quoted = _encodeQuotedPrintableChar(rune, codec);
        wordCounter += quoted.length;
        isWordSplit = wordCounter > qpWordSize;
        if (!isWordSplit) {
          buffer.write(quoted);
        }
      }
      if (isWordSplit) {
        wordCounter = 0;
        runeIndex--;
      }
      if (runeIndex == endIndex) {
        buffer.write(qpWordTail);
      }
    }
    return buffer.toString();
  }
}