dart:convert library.

sdk/lib/convert/line_splitter.dart

+// Copyright (c) 2013, the Dart project authors.  Please see the AUTHORS file
+// for details. All rights reserved. Use of this source code is governed by a
+// BSD-style license that can be found in the LICENSE file.
+
+part of dart.convert;
+
+class LineSplitter extends Converter<String, List<String>> {
+  LineSplitter();
+
+  ChunkedConversionSink<Object> startChunkedConversion(
+      ChunkedConversionSink sink) {
+    return new _LineSplitterSink(outputInterface.adapt(sink));
+  }
+
+  ChunkedConversionInterface get inputInterface =>
+      StringConversionSink.INTERFACE;
+
+  ChunkedConversionInterface get outputInterface =>
+      MultiStringConversionSink.INTERFACE;
+}
+
+class LineFuser extends Converter<List<String>, String> {
+  LineFuser();
+
+  String convert(List<String> input) => input.join("\n");
+
+  fuse(Converter other) => other.fuseInput(this);
+  fuseInput(Converter other) => new FusedConverter(other, this);
+
+  MultiStringConversionSink startChunkedConversion(ChunkedConversionSink sink) {
+    return new _LineFuserSink(outputInterface.adapt(sink));
+  }
+
+  ChunkedConversionInterface get inputInterface =>
+      StringConversionSink.INTERFACE;
+}
+
+class _LineSplitterSink extends StringConversionSinkBase {
+  static const int _LF = 10;
+  static const int _CR = 13;
+
+  Utf8Handler _utf8Handler;
+  MultiStringConversionSink _sink;
+  StringBuffer _carry;
+  // True if the last chunk finished with a _CR.
+  bool _ignoreLineFeed = false;
+
+  _LineSplitterSink(this._sink);
+
+  bool get _hasPartialUtf8Input {
+    return _utf8Handler != null && _utf8Handler.hasPartialInput;
+  }
+
+  bool get _hasPartialInput {
+    if (_carry != null && _carry.isNotEmpty) return true;
+    return _hasPartialUtf8Input;
+  }
+
+  int _findFirstLineSplit(List<int> charCodes, int start, int end) {
+    for (int i = start; i < end; i++) {
+      int codeUnit = charCodes[i];
+      if (codeUnit == _LF || codeUnit == _CR) {
+        return i;
+      };
+    }
+    return end;
+  }
+
+  void _addCarry(List<int> charCodes, int start, int end) {
+    if (_utf8Handler != null && _utf8Handler.hasPartialInput) {
+      throw "Incomplete UTF8 sequence";
+    }
+    if (_carry == null) _carry = new StringBuffer();
+    for (int i = start; i < end; i++) {
+      _carry.writeCharCode(charCodes[i]);
+    }
+  }
+
+  void _addUtf8Carry(List<int> codeUnits, int start, int end) {
+    if (_utf8Handler == null) _utf8Handler = new Utf8Handler();
+    if (_carry == null) _carry = new StringBuffer();
+    _utf8Handler.convertToSink(codeUnits, start, end, _carry);
+  }
+
+  void _split(
+      List<int> charCodes,
+      int startIndex,
+      int endIndex,
+      bool isLast,
+      void addLine(List<int> charCodes, int start, int end, bool isLast),
+      void addCarry(List<int> charCodes, int start, int end)) {
+    int lineStart = startIndex;
+    // If the last chunk ended with a carriage-return we have to ignore a
+    // leading line feed.
+    if (_ignoreLineFeed) {
+      if (startIndex != endIndex && charCodes[startIndex] == _LF) {
+        startIndex++;
+        _ignoreLineFeed = false;
+      }
+    }
+    // Set the `_ignoreLineFeed` boolean for the next chunk.
+    if (startIndex != endIndex) {
+      if (charCodes[endIndex - 1] == _CR) {
+        _ignoreLineFeed = true;
+      }
+    }
+    // If there is a carry we have to append the next part of the line to the
+    // buffer.
+    if (_hasPartialInput) {
+      int firstSplit = _findFirstLineSplit(charCodes, startIndex, endIndex);
+      addCarry(charCodes, startIndex, firstSplit);
+      if (firstSplit == endIndex) {
+        // If `isLast` then nothing comes anymore. Send the carry.
+        if (isLast) {
+          if (_hasPartialUtf8Input) {
+            throw "Incomplete Utf8 sequence";
+          }
+          _sink.addString(_carry.toString(), true);
+        }
+        // Otherwise just return and wait for the next chunk.
+        return;
+      } else {
+        // Send the string that is in the carry and reset it.
+        _sink.addString(_carry.toString(), false);
+        _carry.clear();
+        lineStart = firstSplit + 1;
+        if (charCodes[firstSplit] == _CR &&
+            firstSplit + 1 < endIndex &&
+            charCodes[firstSplit + 1] == _LF) {
+          // Ignore LF following a CR.
+          lineStart++;
+        }
+      }
+    }
+    assert(!_hasPartialInput);
+
+    // At this point the carry is empty. We are now exclusively working on the
+    // new chunk.
+    for (int i = lineStart; i < endIndex; i++) {
+      int codeUnit = charCodes[i];
+      if (codeUnit == _LF) {
+        bool isLastLine = isLast && i == endIndex - 1;
+        addLine(charCodes, lineStart, i, isLastLine);
+        lineStart = i + 1;
+      } else if (codeUnit == _CR) {
+        int lineEnd = i;
+        if (i + 1 < charCodes.length && charCodes[i + 1] == _LF) {
+          i++;
+        }
+        bool isLastLine = isLast && i == endIndex - 1;
+        addLine(charCodes, lineStart, lineEnd, isLastLine);
+        lineStart = i + 1;
+      }
+    }
+    if (lineStart != endIndex) {
+      // The last part of the chunk was not consumed. Store it in the carry or
+      // send it directly if this is the last chunk.
+      if (isLast) {
+        addLine(charCodes, lineStart, endIndex, true);
+      } else {
+        addCarry(charCodes, lineStart, endIndex);
+      }
+    } else if (isLast) {
+      _sink.close();
+    }
+  }
+
+  void add(String str) => addString(str, false);
+
+  void addString(String str, bool isLast) {
+    _split(str.codeUnits, 0, str.length, isLast, _sink.addCodeUnits, _addCarry);
+  }
+
+  void addCodeUnits(List<int> codeUnits, int start, int end, bool isLast) {
+    _split(codeUnits, start, end, isLast, _sink.addCodeUnits, _addCarry);
+  }
+
+  void addRunes(List<int> runes, int start, int end, bool isLast) {
+    _split(runes, start, end, isLast, _sink.addRunes, _addCarry);
+  }
+
+  void addUtf8(List<int> utf8Units, int start, int end, bool isLast) {
+    _split(utf8Units, start, end, isLast, _sink.addUtf8, _addUtf8Carry);
+  }
+
+  void addIsoLatin1(List<int> codeUnits, int start, int end, bool isLast) {
+    _split(codeUnits, start, end, isLast, _sink.addIsoLatin1, _addCarry);
+  }
+
+  void addAscii(List<int> codeUnits, int start, int end, bool isLast) {
+    _split(codeUnits, start, end, isLast, _sink.addAscii, _addCarry);
+  }
+}
+
+class _LineFuserSink implements MultiStringConversionSink {
+  static const _NL = 10;  // New line code.
+  bool needsSeparator = false;
+  StringConversionSink _sink;
+
+  _LineFuserSink(this._sink);
+
+  var _handler = null;
+
+  void writeNewLineIfNecessary() {
+    if (needsSeparator) _sink.writeCharCode(_NL);
+    needsSeparator = true;
+  }
+
+  void addNonChunked(List<String> list) {
+    _sink.writeAll(list, "\n");
+    _sink.close();
+  }
+
+  void add(String str) {
+    writeNewLineIfNecessary();
+    _sink.add(str);
+  }
+
+  void addString(String str, bool isLast) {
+    writeNewLineIfNecessary();
+    _sink.addString(str, isLast);
+  }
+
+  void addCodeUnits(List<int> codeUnits, int start, int end, bool isLast) {
+    writeNewLineIfNecessary();
+    _sink.addCodeUnits(codeUnits, start, end, isLast);
+  }
+  void addRunes(List<int> runes, int start, int end, bool isLast) {
+    writeNewLineIfNecessary();
+    _sink.addRunes(runes, start, end, isLast);
+  }
+  void addUtf8(List<int> utf8Units, int start, int end, bool isLast) {
+    writeNewLineIfNecessary();
+    _sink.addUtf8(utf8Units, start, end, isLast);
+  }
+  void addIsoLatin1(List<int> codeUnits, int start, int end, bool isLast) {
+    writeNewLineIfNecessary();
+    _sink.addIsoLatin1(codeUnits, start, end, isLast);
+  }
+  void addAscii(List<int> codeUnits, int start, int end, bool isLast) {
+    writeNewLineIfNecessary();
+    _sink.addAscii(codeUnits, start, end, isLast);
+  }
+  void iterateCodeUnits(StringIterator iterator, bool isLast) {
+    writeNewLineIfNecessary();
+    _sink.iterateCodeUnits(iterator, isLast);
+  }
+  void iterateRunes(StringIterator iterator, bool isLast) {
+    writeNewLineIfNecessary();
+    _sink.iterateRunes(iterator, isLast);
+  }
+  void iterateUtf8(StringIterator iterator, bool isLast) {
+    writeNewLineIfNecessary();
+    _sink.iterateUtf8(iterator, isLast);
+  }
+  void iterateIsoLatin1(StringIterator iterator, bool isLast) {
+    writeNewLineIfNecessary();
+    _sink.iterateIsoLatin1(iterator, isLast);
+  }
+  void iterateAscii(StringIterator iterator, bool isLast) {
+    writeNewLineIfNecessary();
+    _sink.iterateAscii(iterator, isLast);
+  }
+
+  void write(Object o) {
+    writeNewLineIfNecessary();
+    _sink.write(o);
+  }
+  void writeln([Object o]) {
+    writeNewLineIfNecessary();
+    _sink.writeln(o);
+  }
+  void writeCharCode(int charCode) {
+    writeNewLineIfNecessary();
+    _sink.writeCharCode(charCode);
+  }
+  void writeAll(Iterable objects, [String separator = ""]) {
+    writeNewLineIfNecessary();
+    _sink.writeAll(objects, separator);
+  }
+
+  close() {
+    _sink.close();
+  }
+
+  ChunkedConversionInterface get interface =>
+      MultiStringConversionSink.INTERFACE;
+}