OLD | NEW |
1 // Copyright (c) 2013, the Dart project authors. Please see the AUTHORS file | 1 // Copyright (c) 2013, the Dart project authors. Please see the AUTHORS file |
2 // for details. All rights reserved. Use of this source code is governed by a | 2 // for details. All rights reserved. Use of this source code is governed by a |
3 // BSD-style license that can be found in the LICENSE file. | 3 // BSD-style license that can be found in the LICENSE file. |
4 | 4 |
5 part of dart.io; | 5 part of dart.io; |
6 | |
7 // Interface for decoders decoding binary data into string data. The | |
8 // decoder keeps track of line breaks during decoding. | |
9 abstract class _StringDecoder { | |
10 // Add more binary data to be decoded. The ownership of the buffer | |
11 // is transfered to the decoder and the caller most not modify it any more. | |
12 int write(List<int> buffer); | |
13 | |
14 void done(); | |
15 | |
16 // Returns whether any decoded data is available. | |
17 bool get isEmpty; | |
18 | |
19 // Returns the number of available decoded characters. | |
20 int available(); | |
21 | |
22 // Get the number of line breaks present in the current decoded | |
23 // data. | |
24 int get lineBreaks; | |
25 | |
26 // Get up to [len] characters of string data decoded since the last | |
27 // call to [decode] or [decodeLine]. Returns null if no decoded data | |
28 // is available. If [len] is not specified all decoded characters | |
29 // are returned. | |
30 String decoded([int len]); | |
31 | |
32 // Get the string data decoded since the last call to [decode] or | |
33 // [decodeLine] up to the next line break present. Returns null if | |
34 // no line break is present. The line break character sequence is | |
35 // discarded. | |
36 String get decodedLine; | |
37 | |
38 // Set the handler that will be called if an error ocurs while decoding. | |
39 void set onError(Function callback); | |
40 } | |
41 | |
42 | |
43 class _StringDecoders { | |
44 static _StringDecoder decoder(Encoding encoding) { | |
45 if (encoding == Encoding.UTF_8) { | |
46 return new _UTF8Decoder(); | |
47 } else if (encoding == Encoding.ISO_8859_1) { | |
48 return new _Latin1Decoder(); | |
49 } else if (encoding == Encoding.ASCII) { | |
50 return new _AsciiDecoder(); | |
51 } else if (encoding == Encoding.SYSTEM) { | |
52 if (Platform.operatingSystem == 'windows') { | |
53 return new _WindowsCodePageDecoder(); | |
54 } | |
55 return new _UTF8Decoder(); | |
56 } else { | |
57 if (encoding is Encoding) { | |
58 throw new StreamException("Unsupported encoding ${encoding.name}"); | |
59 } else { | |
60 throw new StreamException("Unsupported encoding ${encoding}"); | |
61 } | |
62 } | |
63 } | |
64 } | |
65 | |
66 | |
67 class DecoderException implements Exception { | |
68 const DecoderException([String this.message]); | |
69 String toString() => "DecoderException: $message"; | |
70 final String message; | |
71 } | |
72 | |
73 | |
74 // Utility class for decoding UTF-8 from data delivered as a stream of | |
75 // bytes. | |
76 abstract class _StringDecoderBase implements _StringDecoder { | |
77 _StringDecoderBase() | |
78 : _bufferList = new _BufferList(), | |
79 _result = new List<int>(), | |
80 _lineBreakEnds = new Queue<int>(); | |
81 | |
82 int write(List<int> buffer) { | |
83 _bufferList.add(buffer); | |
84 // Decode as many bytes into characters as possible. | |
85 while (_bufferList.length > 0) { | |
86 if (!_processNext()) { | |
87 break; | |
88 } | |
89 } | |
90 return buffer.length; | |
91 } | |
92 | |
93 void done() { } | |
94 | |
95 bool get isEmpty => _result.isEmpty; | |
96 | |
97 int get lineBreaks => _lineBreaks; | |
98 | |
99 String decoded([int length]) { | |
100 if (isEmpty) return null; | |
101 | |
102 List<int> result; | |
103 if (length != null && length < available()) { | |
104 result = _result.getRange(_resultOffset, length); | |
105 } else if (_resultOffset == 0) { | |
106 result = _result; | |
107 } else { | |
108 result = _result.getRange(_resultOffset, _result.length - _resultOffset); | |
109 } | |
110 | |
111 _resultOffset += result.length; | |
112 while (!_lineBreakEnds.isEmpty && | |
113 _lineBreakEnds.first < _charOffset + _resultOffset) { | |
114 _lineBreakEnds.removeFirst(); | |
115 _lineBreaks--; | |
116 } | |
117 if (_result.length == _resultOffset) _resetResult(); | |
118 return new String.fromCharCodes(result); | |
119 } | |
120 | |
121 String get decodedLine { | |
122 if (_lineBreakEnds.isEmpty) return null; | |
123 int lineEnd = _lineBreakEnds.removeFirst(); | |
124 int terminationSequenceLength = 1; | |
125 if (_result[lineEnd - _charOffset] == LF && | |
126 lineEnd > _charOffset && | |
127 _resultOffset < lineEnd && | |
128 _result[lineEnd - _charOffset - 1] == CR) { | |
129 terminationSequenceLength = 2; | |
130 } | |
131 var lineLength = | |
132 lineEnd - _charOffset - _resultOffset - terminationSequenceLength + 1; | |
133 String result = | |
134 new String.fromCharCodes(_result.getRange(_resultOffset, lineLength)); | |
135 _lineBreaks--; | |
136 _resultOffset += (lineLength + terminationSequenceLength); | |
137 if (_result.length == _resultOffset) _resetResult(); | |
138 return result; | |
139 } | |
140 | |
141 // Add another decoded character. | |
142 void addChar(int charCode) { | |
143 _result.add(charCode); | |
144 _charCount++; | |
145 // Check for line ends (\r, \n and \r\n). | |
146 if (charCode == LF) { | |
147 _recordLineBreakEnd(_charCount - 1); | |
148 } else if (_lastCharCode == CR) { | |
149 _recordLineBreakEnd(_charCount - 2); | |
150 } | |
151 _lastCharCode = charCode; | |
152 } | |
153 | |
154 int available() => _result.length - _resultOffset; | |
155 | |
156 void _recordLineBreakEnd(int charPos) { | |
157 _lineBreakEnds.add(charPos); | |
158 _lineBreaks++; | |
159 } | |
160 | |
161 void _resetResult() { | |
162 _charOffset += _result.length; | |
163 _result = new List<int>(); | |
164 _resultOffset = 0; | |
165 } | |
166 | |
167 bool _processNext(); | |
168 | |
169 _BufferList _bufferList; | |
170 int _resultOffset = 0; | |
171 List<int> _result; | |
172 int _lineBreaks = 0; // Number of line breaks in the current list. | |
173 // The positions of the line breaks are tracked in terms of absolute | |
174 // character positions from the begining of the decoded data. | |
175 Queue<int> _lineBreakEnds; // Character position of known line breaks. | |
176 int _charOffset = 0; // Character number of the first character in the list. | |
177 int _charCount = 0; // Total number of characters decoded. | |
178 int _lastCharCode = -1; | |
179 Function onError; | |
180 | |
181 final int LF = 10; | |
182 final int CR = 13; | |
183 } | |
184 | |
185 | |
186 // Utility class for decoding UTF-8 from data delivered as a stream of | |
187 // bytes. | |
188 class _UTF8Decoder extends _StringDecoderBase { | |
189 static const kMaxCodePoint = 0x10FFFF; | |
190 static const kReplacementCodePoint = 0x3f; | |
191 | |
192 void done() { | |
193 if (!_bufferList.isEmpty) { | |
194 _reportError(new DecoderException("Illegal UTF-8")); | |
195 } | |
196 } | |
197 | |
198 bool _reportError(error) { | |
199 if (onError != null) { | |
200 onError(error); | |
201 return false; | |
202 } else { | |
203 throw error; | |
204 } | |
205 } | |
206 | |
207 // Process the next UTF-8 encoded character. | |
208 bool _processNext() { | |
209 // Peek the next byte to calculate the number of bytes required for | |
210 // the next character. | |
211 int value = _bufferList.peek() & 0xFF; | |
212 if ((value & 0x80) == 0x80) { | |
213 int additionalBytes; | |
214 if ((value & 0xe0) == 0xc0) { // 110xxxxx | |
215 value = value & 0x1F; | |
216 additionalBytes = 1; | |
217 } else if ((value & 0xf0) == 0xe0) { // 1110xxxx | |
218 value = value & 0x0F; | |
219 additionalBytes = 2; | |
220 } else if ((value & 0xf8) == 0xf0) { // 11110xxx | |
221 value = value & 0x07; | |
222 additionalBytes = 3; | |
223 } else if ((value & 0xfc) == 0xf8) { // 111110xx | |
224 value = value & 0x03; | |
225 additionalBytes = 4; | |
226 } else if ((value & 0xfe) == 0xfc) { // 1111110x | |
227 value = value & 0x01; | |
228 additionalBytes = 5; | |
229 } else { | |
230 return _reportError(new DecoderException("Illegal UTF-8")); | |
231 } | |
232 // Check if there are enough bytes to decode the character. Otherwise | |
233 // return false. | |
234 if (_bufferList.length < additionalBytes + 1) { | |
235 return false; | |
236 } | |
237 // Remove the value peeked from the buffer list. | |
238 _bufferList.next(); | |
239 for (int i = 0; i < additionalBytes; i++) { | |
240 int byte = _bufferList.next(); | |
241 if ((byte & 0xc0) != 0x80) { | |
242 return _reportError(new DecoderException("Illegal UTF-8")); | |
243 } | |
244 value = value << 6 | (byte & 0x3F); | |
245 } | |
246 } else { | |
247 // Remove the value peeked from the buffer list. | |
248 _bufferList.next(); | |
249 } | |
250 if (value > kMaxCodePoint) { | |
251 addChar(kReplacementCodePoint); | |
252 } else { | |
253 addChar(value); | |
254 } | |
255 return true; | |
256 } | |
257 } | |
258 | |
259 | |
260 // Utility class for decoding ascii data delivered as a stream of | |
261 // bytes. | |
262 class _AsciiDecoder extends _StringDecoderBase { | |
263 // Process the next ascii encoded character. | |
264 bool _processNext() { | |
265 while (_bufferList.length > 0) { | |
266 int byte = _bufferList.next(); | |
267 if (byte > 127) { | |
268 var error = new DecoderException("Illegal ASCII character $byte"); | |
269 if (onError != null) { | |
270 onError(error); | |
271 return false; | |
272 } else { | |
273 throw error; | |
274 } | |
275 } | |
276 addChar(byte); | |
277 } | |
278 return true; | |
279 } | |
280 } | |
281 | |
282 | |
283 // Utility class for decoding Latin-1 data delivered as a stream of | |
284 // bytes. | |
285 class _Latin1Decoder extends _StringDecoderBase { | |
286 // Process the next Latin-1 encoded character. | |
287 bool _processNext() { | |
288 while (_bufferList.length > 0) { | |
289 int byte = _bufferList.next(); | |
290 addChar(byte); | |
291 } | |
292 return true; | |
293 } | |
294 } | |
295 | |
296 | |
297 // Utility class for decoding Windows current code page data delivered | |
298 // as a stream of bytes. | |
299 class _WindowsCodePageDecoder extends _StringDecoderBase { | |
300 // Process the next chunk of data. | |
301 bool _processNext() { | |
302 List<int> bytes = _bufferList.readBytes(_bufferList.length); | |
303 for (var charCode in _decodeBytes(bytes).charCodes) { | |
304 addChar(charCode); | |
305 } | |
306 return true; | |
307 } | |
308 | |
309 external static String _decodeBytes(List<int> bytes); | |
310 } | |
311 | |
312 | |
313 // Interface for encoders encoding string data into binary data. | |
314 abstract class _StringEncoder { | |
315 List<int> encodeString(String string); | |
316 } | |
317 | |
318 | |
319 // Utility class for encoding a string into UTF-8 byte stream. | |
320 class _UTF8Encoder implements _StringEncoder { | |
321 List<int> encodeString(String string) { | |
322 int size = _encodingSize(string); | |
323 List<int> result = new Uint8List(size); | |
324 _encodeString(string, result); | |
325 return result; | |
326 } | |
327 | |
328 static int _encodingSize(String string) => _encodeString(string, null); | |
329 | |
330 static int _encodeString(String string, List<int> buffer) { | |
331 List<int> utf8CodeUnits = encodeUtf8(string); | |
332 if (buffer != null) { | |
333 for (int i = 0; i < utf8CodeUnits.length; i++) { | |
334 buffer[i] = utf8CodeUnits[i]; | |
335 } | |
336 } | |
337 return utf8CodeUnits.length; | |
338 } | |
339 } | |
340 | |
341 | |
342 // Utility class for encoding a string into a Latin1 byte stream. | |
343 class _Latin1Encoder implements _StringEncoder { | |
344 List<int> encodeString(String string) { | |
345 List<int> result = new Uint8List(string.length); | |
346 for (int i = 0; i < string.length; i++) { | |
347 int charCode = string.charCodeAt(i); | |
348 if (charCode > 255) { | |
349 throw new EncoderException( | |
350 "No ISO_8859_1 encoding for code point $charCode"); | |
351 } | |
352 result[i] = charCode; | |
353 } | |
354 return result; | |
355 } | |
356 } | |
357 | |
358 | |
359 // Utility class for encoding a string into an ASCII byte stream. | |
360 class _AsciiEncoder implements _StringEncoder { | |
361 List<int> encodeString(String string) { | |
362 List<int> result = new Uint8List(string.length); | |
363 for (int i = 0; i < string.length; i++) { | |
364 int charCode = string.charCodeAt(i); | |
365 if (charCode > 127) { | |
366 throw new EncoderException( | |
367 "No ASCII encoding for code point $charCode"); | |
368 } | |
369 result[i] = charCode; | |
370 } | |
371 return result; | |
372 } | |
373 } | |
374 | |
375 | |
376 // Utility class for encoding a string into a current windows | |
377 // code page byte list. | |
378 class _WindowsCodePageEncoder implements _StringEncoder { | |
379 List<int> encodeString(String string) { | |
380 return _encodeString(string); | |
381 } | |
382 | |
383 external static List<int> _encodeString(String string); | |
384 } | |
385 | |
386 | |
387 class _StringEncoders { | |
388 static _StringEncoder encoder(Encoding encoding) { | |
389 if (encoding == Encoding.UTF_8) { | |
390 return new _UTF8Encoder(); | |
391 } else if (encoding == Encoding.ISO_8859_1) { | |
392 return new _Latin1Encoder(); | |
393 } else if (encoding == Encoding.ASCII) { | |
394 return new _AsciiEncoder(); | |
395 } else if (encoding == Encoding.SYSTEM) { | |
396 if (Platform.operatingSystem == 'windows') { | |
397 return new _WindowsCodePageEncoder(); | |
398 } | |
399 return new _UTF8Encoder(); | |
400 } else { | |
401 throw new StreamException("Unsupported encoding ${encoding.name}"); | |
402 } | |
403 } | |
404 } | |
405 | |
406 | |
407 class EncoderException implements Exception { | |
408 const EncoderException([String this.message]); | |
409 String toString() => "EncoderException: $message"; | |
410 final String message; | |
411 } | |
412 | |
413 | |
414 class _StringInputStream implements StringInputStream { | |
415 _StringInputStream(InputStream this._input, Encoding this._encoding) { | |
416 _decoder = _StringDecoders.decoder(encoding); | |
417 _input.onData = _onData; | |
418 _input.onClosed = _onClosed; | |
419 } | |
420 | |
421 String read([int len]) { | |
422 String result = _decoder.decoded(len); | |
423 _checkInstallDataHandler(); | |
424 return result; | |
425 } | |
426 | |
427 String readLine() { | |
428 String decodedLine = _decoder.decodedLine; | |
429 if (decodedLine == null) { | |
430 if (_inputClosed) { | |
431 // Last line might not have a line separator. | |
432 decodedLine = _decoder.decoded(); | |
433 if (decodedLine != null && | |
434 decodedLine[decodedLine.length - 1] == '\r') { | |
435 decodedLine = decodedLine.substring(0, decodedLine.length - 1); | |
436 } | |
437 } | |
438 } | |
439 _checkInstallDataHandler(); | |
440 return decodedLine; | |
441 } | |
442 | |
443 int available() => _decoder.available(); | |
444 | |
445 Encoding get encoding => _encoding; | |
446 | |
447 bool get closed => _inputClosed && _decoder.isEmpty; | |
448 | |
449 void set onData(void callback()) { | |
450 _clientDataHandler = callback; | |
451 _clientLineHandler = null; | |
452 _checkInstallDataHandler(); | |
453 _checkScheduleCallback(); | |
454 } | |
455 | |
456 void set onLine(void callback()) { | |
457 _clientLineHandler = callback; | |
458 _clientDataHandler = null; | |
459 _checkInstallDataHandler(); | |
460 _checkScheduleCallback(); | |
461 } | |
462 | |
463 void set onClosed(void callback()) { | |
464 _clientCloseHandler = callback; | |
465 } | |
466 | |
467 void set onError(void callback(e)) { | |
468 _input.onError = callback; | |
469 _decoder.onError = (e) { | |
470 _clientCloseHandler = null; | |
471 _input.close(); | |
472 callback(e); | |
473 }; | |
474 } | |
475 | |
476 void _onData() { | |
477 _readData(); | |
478 if (!_decoder.isEmpty && _clientDataHandler != null) { | |
479 _clientDataHandler(); | |
480 } | |
481 if (_decoder.lineBreaks > 0 && _clientLineHandler != null) { | |
482 _clientLineHandler(); | |
483 } | |
484 _checkScheduleCallback(); | |
485 _checkInstallDataHandler(); | |
486 } | |
487 | |
488 void _onClosed() { | |
489 _inputClosed = true; | |
490 _decoder.done(); | |
491 if (_decoder.isEmpty && _clientCloseHandler != null) { | |
492 _clientCloseHandler(); | |
493 } else { | |
494 _checkScheduleCallback(); | |
495 } | |
496 } | |
497 | |
498 void _readData() { | |
499 List<int> data = _input.read(); | |
500 if (data != null) { | |
501 _decoder.write(data); | |
502 } | |
503 } | |
504 | |
505 void _checkInstallDataHandler() { | |
506 if (_inputClosed || | |
507 (_clientDataHandler == null && _clientLineHandler == null)) { | |
508 _input.onData = null; | |
509 } else if (_clientDataHandler != null) { | |
510 if (_decoder.isEmpty) { | |
511 _input.onData = _onData; | |
512 } else { | |
513 _input.onData = null; | |
514 } | |
515 } else { | |
516 assert(_clientLineHandler != null); | |
517 if (_decoder.lineBreaks == 0) { | |
518 _input.onData = _onData; | |
519 } else { | |
520 _input.onData = null; | |
521 } | |
522 } | |
523 } | |
524 | |
525 // TODO(sgjesse): Find a better way of scheduling callbacks from | |
526 // the event loop. | |
527 void _checkScheduleCallback() { | |
528 void issueDataCallback() { | |
529 _scheduledDataCallback = null; | |
530 if (_clientDataHandler != null) { | |
531 _clientDataHandler(); | |
532 _checkScheduleCallback(); | |
533 } | |
534 } | |
535 | |
536 void issueLineCallback() { | |
537 _scheduledLineCallback = null; | |
538 if (_clientLineHandler != null) { | |
539 _clientLineHandler(); | |
540 _checkScheduleCallback(); | |
541 } | |
542 } | |
543 | |
544 void issueCloseCallback() { | |
545 _scheduledCloseCallback = null; | |
546 if (!_closed) { | |
547 if (_clientCloseHandler != null) _clientCloseHandler(); | |
548 _closed = true; | |
549 } | |
550 } | |
551 | |
552 if (!_closed) { | |
553 // Schedule data callback if string data available. | |
554 if (_clientDataHandler != null && | |
555 !_decoder.isEmpty && | |
556 _scheduledDataCallback == null) { | |
557 if (_scheduledLineCallback != null) { | |
558 _scheduledLineCallback.cancel(); | |
559 } | |
560 _scheduledDataCallback = Timer.run(issueDataCallback); | |
561 } | |
562 | |
563 // Schedule line callback if a line is available. | |
564 if (_clientLineHandler != null && | |
565 (_decoder.lineBreaks > 0 || (!_decoder.isEmpty && _inputClosed)) && | |
566 _scheduledLineCallback == null) { | |
567 if (_scheduledDataCallback != null) { | |
568 _scheduledDataCallback.cancel(); | |
569 } | |
570 _scheduledLineCallback = Timer.run(issueLineCallback); | |
571 } | |
572 | |
573 // Schedule close callback if no more data and input is closed. | |
574 if (_decoder.isEmpty && | |
575 _inputClosed && | |
576 _scheduledCloseCallback == null) { | |
577 _scheduledCloseCallback = Timer.run(issueCloseCallback); | |
578 } | |
579 } | |
580 } | |
581 | |
582 InputStream _input; | |
583 Encoding _encoding; | |
584 _StringDecoder _decoder; | |
585 bool _inputClosed = false; // Is the underlying input stream closed? | |
586 bool _closed = false; // Is this stream closed. | |
587 bool _eof = false; // Has all data been read from the decoder? | |
588 Timer _scheduledDataCallback; | |
589 Timer _scheduledLineCallback; | |
590 Timer _scheduledCloseCallback; | |
591 Function _clientDataHandler; | |
592 Function _clientLineHandler; | |
593 Function _clientCloseHandler; | |
594 } | |
OLD | NEW |