OLD | NEW |
| (Empty) |
1 // Copyright (c) 2014, the Dart project authors. Please see the AUTHORS file | |
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. | |
4 | |
5 // This code was auto-generated, is not intended to be edited, and is subject to | |
6 // significant change. Please see the README file for more information. | |
7 | |
8 library engine.html; | |
9 | |
10 import 'dart:collection'; | |
11 | |
12 import 'ast.dart'; | |
13 import 'element.dart'; | |
14 import 'engine.dart' show AnalysisOptions, AnalysisEngine; | |
15 import 'error.dart' show AnalysisErrorListener; | |
16 import 'java_core.dart'; | |
17 import 'java_engine.dart'; | |
18 import 'parser.dart' show Parser; | |
19 import 'scanner.dart' as sc show Scanner, SubSequenceReader, Token; | |
20 import 'source.dart'; | |
21 | |
22 /** | |
23 * The abstract class `AbstractScanner` implements a scanner for HTML code. Subc
lasses are | |
24 * required to implement the interface used to access the characters being scann
ed. | |
25 */ | |
26 abstract class AbstractScanner { | |
27 static List<String> _NO_PASS_THROUGH_ELEMENTS = <String>[]; | |
28 | |
29 /** | |
30 * The source being scanned. | |
31 */ | |
32 final Source source; | |
33 | |
34 /** | |
35 * The token pointing to the head of the linked list of tokens. | |
36 */ | |
37 Token _tokens; | |
38 | |
39 /** | |
40 * The last token that was scanned. | |
41 */ | |
42 Token _tail; | |
43 | |
44 /** | |
45 * A list containing the offsets of the first character of each line in the so
urce code. | |
46 */ | |
47 List<int> _lineStarts = new List<int>(); | |
48 | |
49 /** | |
50 * An array of element tags for which the content between tags should be consi
der a single token. | |
51 */ | |
52 List<String> _passThroughElements = _NO_PASS_THROUGH_ELEMENTS; | |
53 | |
54 /** | |
55 * Initialize a newly created scanner. | |
56 * | |
57 * @param source the source being scanned | |
58 */ | |
59 AbstractScanner(this.source) { | |
60 _tokens = new Token.con1(TokenType.EOF, -1); | |
61 _tokens.setNext(_tokens); | |
62 _tail = _tokens; | |
63 recordStartOfLine(); | |
64 } | |
65 | |
66 /** | |
67 * Return an array containing the offsets of the first character of each line
in the source code. | |
68 * | |
69 * @return an array containing the offsets of the first character of each line
in the source code | |
70 */ | |
71 List<int> get lineStarts => _lineStarts; | |
72 | |
73 /** | |
74 * Return the current offset relative to the beginning of the file. Return the
initial offset if | |
75 * the scanner has not yet scanned the source code, and one (1) past the end o
f the source code if | |
76 * the source code has been scanned. | |
77 * | |
78 * @return the current offset of the scanner in the source | |
79 */ | |
80 int get offset; | |
81 | |
82 /** | |
83 * Set array of element tags for which the content between tags should be cons
ider a single token. | |
84 */ | |
85 void set passThroughElements(List<String> passThroughElements) { | |
86 this._passThroughElements = passThroughElements != null | |
87 ? passThroughElements | |
88 : _NO_PASS_THROUGH_ELEMENTS; | |
89 } | |
90 | |
91 /** | |
92 * Advance the current position and return the character at the new current po
sition. | |
93 * | |
94 * @return the character at the new current position | |
95 */ | |
96 int advance(); | |
97 | |
98 /** | |
99 * Return the substring of the source code between the start offset and the mo
dified current | |
100 * position. The current position is modified by adding the end delta. | |
101 * | |
102 * @param start the offset to the beginning of the string, relative to the sta
rt of the file | |
103 * @param endDelta the number of character after the current location to be in
cluded in the | |
104 * string, or the number of characters before the current location to
be excluded if the | |
105 * offset is negative | |
106 * @return the specified substring of the source code | |
107 */ | |
108 String getString(int start, int endDelta); | |
109 | |
110 /** | |
111 * Return the character at the current position without changing the current p
osition. | |
112 * | |
113 * @return the character at the current position | |
114 */ | |
115 int peek(); | |
116 | |
117 /** | |
118 * Record the fact that we are at the beginning of a new line in the source. | |
119 */ | |
120 void recordStartOfLine() { | |
121 _lineStarts.add(offset); | |
122 } | |
123 | |
124 /** | |
125 * Scan the source code to produce a list of tokens representing the source. | |
126 * | |
127 * @return the first token in the list of tokens that were produced | |
128 */ | |
129 Token tokenize() { | |
130 _scan(); | |
131 _appendEofToken(); | |
132 return _firstToken(); | |
133 } | |
134 | |
135 void _appendEofToken() { | |
136 Token eofToken = new Token.con1(TokenType.EOF, offset); | |
137 // The EOF token points to itself so that there is always infinite | |
138 // look-ahead. | |
139 eofToken.setNext(eofToken); | |
140 _tail = _tail.setNext(eofToken); | |
141 } | |
142 | |
143 Token _emit(Token token) { | |
144 _tail.setNext(token); | |
145 _tail = token; | |
146 return token; | |
147 } | |
148 | |
149 Token _emitWithOffset(TokenType type, int start) => | |
150 _emit(new Token.con1(type, start)); | |
151 | |
152 Token _emitWithOffsetAndLength(TokenType type, int start, int count) => | |
153 _emit(new Token.con2(type, start, getString(start, count))); | |
154 | |
155 Token _firstToken() => _tokens.next; | |
156 | |
157 int _recordStartOfLineAndAdvance(int c) { | |
158 if (c == 0xD) { | |
159 c = advance(); | |
160 if (c == 0xA) { | |
161 c = advance(); | |
162 } | |
163 recordStartOfLine(); | |
164 } else if (c == 0xA) { | |
165 c = advance(); | |
166 recordStartOfLine(); | |
167 } else { | |
168 c = advance(); | |
169 } | |
170 return c; | |
171 } | |
172 | |
173 void _scan() { | |
174 bool inBrackets = false; | |
175 String endPassThrough = null; | |
176 int c = advance(); | |
177 while (c >= 0) { | |
178 int start = offset; | |
179 if (c == 0x3C) { | |
180 c = advance(); | |
181 if (c == 0x21) { | |
182 c = advance(); | |
183 if (c == 0x2D && peek() == 0x2D) { | |
184 // handle a comment | |
185 c = advance(); | |
186 int dashCount = 1; | |
187 while (c >= 0) { | |
188 if (c == 0x2D) { | |
189 dashCount++; | |
190 } else if (c == 0x3E && dashCount >= 2) { | |
191 c = advance(); | |
192 break; | |
193 } else { | |
194 dashCount = 0; | |
195 } | |
196 c = _recordStartOfLineAndAdvance(c); | |
197 } | |
198 _emitWithOffsetAndLength(TokenType.COMMENT, start, -1); | |
199 // Capture <!--> and <!---> as tokens but report an error | |
200 if (_tail.length < 7) { | |
201 // TODO (danrubel): Report invalid HTML comment | |
202 } | |
203 } else { | |
204 // handle a declaration | |
205 while (c >= 0) { | |
206 if (c == 0x3E) { | |
207 c = advance(); | |
208 break; | |
209 } | |
210 c = _recordStartOfLineAndAdvance(c); | |
211 } | |
212 _emitWithOffsetAndLength(TokenType.DECLARATION, start, -1); | |
213 if (!StringUtilities.endsWithChar(_tail.lexeme, 0x3E)) { | |
214 // TODO (danrubel): Report missing '>' in directive | |
215 } | |
216 } | |
217 } else if (c == 0x3F) { | |
218 // handle a directive | |
219 while (c >= 0) { | |
220 if (c == 0x3F) { | |
221 c = advance(); | |
222 if (c == 0x3E) { | |
223 c = advance(); | |
224 break; | |
225 } | |
226 } else { | |
227 c = _recordStartOfLineAndAdvance(c); | |
228 } | |
229 } | |
230 _emitWithOffsetAndLength(TokenType.DIRECTIVE, start, -1); | |
231 if (_tail.length < 4) { | |
232 // TODO (danrubel): Report invalid directive | |
233 } | |
234 } else if (c == 0x2F) { | |
235 _emitWithOffset(TokenType.LT_SLASH, start); | |
236 inBrackets = true; | |
237 c = advance(); | |
238 } else { | |
239 inBrackets = true; | |
240 _emitWithOffset(TokenType.LT, start); | |
241 // ignore whitespace in braces | |
242 while (Character.isWhitespace(c)) { | |
243 c = _recordStartOfLineAndAdvance(c); | |
244 } | |
245 // get tag | |
246 if (Character.isLetterOrDigit(c)) { | |
247 int tagStart = offset; | |
248 c = advance(); | |
249 while (Character.isLetterOrDigit(c) || c == 0x2D || c == 0x5F) { | |
250 c = advance(); | |
251 } | |
252 _emitWithOffsetAndLength(TokenType.TAG, tagStart, -1); | |
253 // check tag against passThrough elements | |
254 String tag = _tail.lexeme; | |
255 for (String str in _passThroughElements) { | |
256 if (str == tag) { | |
257 endPassThrough = "</$str>"; | |
258 break; | |
259 } | |
260 } | |
261 } | |
262 } | |
263 } else if (c == 0x3E) { | |
264 _emitWithOffset(TokenType.GT, start); | |
265 inBrackets = false; | |
266 c = advance(); | |
267 // if passThrough != null, read until we match it | |
268 if (endPassThrough != null) { | |
269 bool endFound = false; | |
270 int len = endPassThrough.length; | |
271 int firstC = endPassThrough.codeUnitAt(0); | |
272 int index = 0; | |
273 int nextC = firstC; | |
274 while (c >= 0) { | |
275 if (c == nextC) { | |
276 index++; | |
277 if (index == len) { | |
278 endFound = true; | |
279 break; | |
280 } | |
281 nextC = endPassThrough.codeUnitAt(index); | |
282 } else if (c == firstC) { | |
283 index = 1; | |
284 nextC = endPassThrough.codeUnitAt(1); | |
285 } else { | |
286 index = 0; | |
287 nextC = firstC; | |
288 } | |
289 c = _recordStartOfLineAndAdvance(c); | |
290 } | |
291 if (start + 1 < offset) { | |
292 if (endFound) { | |
293 _emitWithOffsetAndLength(TokenType.TEXT, start + 1, -len); | |
294 _emitWithOffset(TokenType.LT_SLASH, offset - len + 1); | |
295 _emitWithOffsetAndLength(TokenType.TAG, offset - len + 3, -1); | |
296 } else { | |
297 _emitWithOffsetAndLength(TokenType.TEXT, start + 1, -1); | |
298 } | |
299 } | |
300 endPassThrough = null; | |
301 } | |
302 } else if (c == 0x2F && peek() == 0x3E) { | |
303 advance(); | |
304 _emitWithOffset(TokenType.SLASH_GT, start); | |
305 inBrackets = false; | |
306 c = advance(); | |
307 } else if (!inBrackets) { | |
308 c = _recordStartOfLineAndAdvance(c); | |
309 while (c != 0x3C && c >= 0) { | |
310 c = _recordStartOfLineAndAdvance(c); | |
311 } | |
312 _emitWithOffsetAndLength(TokenType.TEXT, start, -1); | |
313 } else if (c == 0x22 || c == 0x27) { | |
314 // read a string | |
315 int endQuote = c; | |
316 c = advance(); | |
317 while (c >= 0) { | |
318 if (c == endQuote) { | |
319 c = advance(); | |
320 break; | |
321 } | |
322 c = _recordStartOfLineAndAdvance(c); | |
323 } | |
324 _emitWithOffsetAndLength(TokenType.STRING, start, -1); | |
325 } else if (c == 0x3D) { | |
326 // a non-char token | |
327 _emitWithOffset(TokenType.EQ, start); | |
328 c = advance(); | |
329 } else if (Character.isWhitespace(c)) { | |
330 // ignore whitespace in braces | |
331 do { | |
332 c = _recordStartOfLineAndAdvance(c); | |
333 } while (Character.isWhitespace(c)); | |
334 } else if (Character.isLetterOrDigit(c)) { | |
335 c = advance(); | |
336 while (Character.isLetterOrDigit(c) || c == 0x2D || c == 0x5F) { | |
337 c = advance(); | |
338 } | |
339 _emitWithOffsetAndLength(TokenType.TAG, start, -1); | |
340 } else { | |
341 // a non-char token | |
342 _emitWithOffsetAndLength(TokenType.TEXT, start, 0); | |
343 c = advance(); | |
344 } | |
345 } | |
346 } | |
347 } | |
348 | |
349 /** | |
350 * Instances of the class `HtmlParser` are used to parse tokens into a AST struc
ture comprised | |
351 * of [XmlNode]s. | |
352 */ | |
353 class HtmlParser extends XmlParser { | |
354 static String _APPLICATION_DART_IN_DOUBLE_QUOTES = "\"application/dart\""; | |
355 | |
356 static String _APPLICATION_DART_IN_SINGLE_QUOTES = "'application/dart'"; | |
357 | |
358 static String _SCRIPT = "script"; | |
359 | |
360 static String _TYPE = "type"; | |
361 | |
362 /** | |
363 * A set containing the names of tags that do not have a closing tag. | |
364 */ | |
365 static Set<String> SELF_CLOSING = new HashSet<String>.from(<String>[ | |
366 "area", | |
367 "base", | |
368 "basefont", | |
369 "br", | |
370 "col", | |
371 "frame", | |
372 "hr", | |
373 "img", | |
374 "input", | |
375 "link", | |
376 "meta", | |
377 "param", | |
378 "!" | |
379 ]); | |
380 | |
381 /** | |
382 * The line information associated with the source being parsed. | |
383 */ | |
384 LineInfo _lineInfo; | |
385 | |
386 /** | |
387 * The error listener to which errors will be reported. | |
388 */ | |
389 final AnalysisErrorListener _errorListener; | |
390 | |
391 final AnalysisOptions _options; | |
392 | |
393 /** | |
394 * Construct a parser for the specified source. | |
395 * | |
396 * [source] is the source being parsed. [_errorListener] is the error | |
397 * listener to which errors will be reported. [_options] is the analysis | |
398 * options which should be used for parsing. | |
399 */ | |
400 HtmlParser(Source source, this._errorListener, this._options) : super(source); | |
401 | |
402 @override | |
403 XmlAttributeNode createAttributeNode(Token name, Token equals, Token value) => | |
404 new XmlAttributeNode(name, equals, value); | |
405 | |
406 @override | |
407 XmlTagNode createTagNode(Token nodeStart, Token tag, | |
408 List<XmlAttributeNode> attributes, Token attributeEnd, | |
409 List<XmlTagNode> tagNodes, Token contentEnd, Token closingTag, | |
410 Token nodeEnd) { | |
411 if (_isScriptNode(tag, attributes, tagNodes)) { | |
412 HtmlScriptTagNode tagNode = new HtmlScriptTagNode(nodeStart, tag, | |
413 attributes, attributeEnd, tagNodes, contentEnd, closingTag, nodeEnd); | |
414 String contents = tagNode.content; | |
415 int contentOffset = attributeEnd.end; | |
416 LineInfo_Location location = _lineInfo.getLocation(contentOffset); | |
417 sc.Scanner scanner = new sc.Scanner(source, | |
418 new sc.SubSequenceReader(contents, contentOffset), _errorListener); | |
419 scanner.setSourceStart(location.lineNumber, location.columnNumber); | |
420 sc.Token firstToken = scanner.tokenize(); | |
421 Parser parser = new Parser(source, _errorListener); | |
422 CompilationUnit unit = parser.parseCompilationUnit(firstToken); | |
423 unit.lineInfo = _lineInfo; | |
424 tagNode.script = unit; | |
425 return tagNode; | |
426 } | |
427 return new XmlTagNode(nodeStart, tag, attributes, attributeEnd, tagNodes, | |
428 contentEnd, closingTag, nodeEnd); | |
429 } | |
430 | |
431 @override | |
432 bool isSelfClosing(Token tag) => SELF_CLOSING.contains(tag.lexeme); | |
433 | |
434 /** | |
435 * Parse the given tokens. | |
436 * | |
437 * @param token the first token in the stream of tokens to be parsed | |
438 * @param lineInfo the line information created by the scanner | |
439 * @return the parse result (not `null`) | |
440 */ | |
441 HtmlUnit parse(Token token, LineInfo lineInfo) { | |
442 this._lineInfo = lineInfo; | |
443 List<XmlTagNode> tagNodes = parseTopTagNodes(token); | |
444 return new HtmlUnit(token, tagNodes, currentToken); | |
445 } | |
446 | |
447 /** | |
448 * Determine if the specified node is a Dart script. | |
449 * | |
450 * @param node the node to be tested (not `null`) | |
451 * @return `true` if the node is a Dart script | |
452 */ | |
453 bool _isScriptNode( | |
454 Token tag, List<XmlAttributeNode> attributes, List<XmlTagNode> tagNodes) { | |
455 if (tagNodes.length != 0 || tag.lexeme != _SCRIPT) { | |
456 return false; | |
457 } | |
458 for (XmlAttributeNode attribute in attributes) { | |
459 if (attribute.name == _TYPE) { | |
460 Token valueToken = attribute.valueToken; | |
461 if (valueToken != null) { | |
462 String value = valueToken.lexeme; | |
463 if (value == _APPLICATION_DART_IN_DOUBLE_QUOTES || | |
464 value == _APPLICATION_DART_IN_SINGLE_QUOTES) { | |
465 return true; | |
466 } | |
467 } | |
468 } | |
469 } | |
470 return false; | |
471 } | |
472 | |
473 /** | |
474 * Given the contents of an embedded expression that occurs at the given offse
t, parse it as a | |
475 * Dart expression. The contents should not include the expression's delimiter
s. | |
476 * | |
477 * @param source the source that contains that given token | |
478 * @param token the token to start parsing from | |
479 * @return the Dart expression that was parsed | |
480 */ | |
481 static Expression parseEmbeddedExpression( | |
482 Source source, sc.Token token, AnalysisErrorListener errorListener) { | |
483 Parser parser = new Parser(source, errorListener); | |
484 return parser.parseExpression(token); | |
485 } | |
486 | |
487 /** | |
488 * Given the contents of an embedded expression that occurs at the given offse
t, scans it as a | |
489 * Dart code. | |
490 * | |
491 * @param source the source of that contains the given contents | |
492 * @param contents the contents to scan | |
493 * @param contentOffset the offset of the contents in the larger file | |
494 * @return the first Dart token | |
495 */ | |
496 static sc.Token scanDartSource(Source source, LineInfo lineInfo, | |
497 String contents, int contentOffset, AnalysisErrorListener errorListener) { | |
498 LineInfo_Location location = lineInfo.getLocation(contentOffset); | |
499 sc.Scanner scanner = new sc.Scanner(source, | |
500 new sc.SubSequenceReader(contents, contentOffset), errorListener); | |
501 scanner.setSourceStart(location.lineNumber, location.columnNumber); | |
502 return scanner.tokenize(); | |
503 } | |
504 } | |
505 | |
506 /** | |
507 * Instances of the class `HtmlScriptTagNode` represent a script tag within an H
TML file that | |
508 * references a Dart script. | |
509 */ | |
510 @deprecated | |
511 class HtmlScriptTagNode extends XmlTagNode { | |
512 /** | |
513 * The AST structure representing the Dart code within this tag. | |
514 */ | |
515 CompilationUnit _script; | |
516 | |
517 /** | |
518 * The element representing this script. | |
519 */ | |
520 HtmlScriptElement scriptElement; | |
521 | |
522 /** | |
523 * Initialize a newly created node to represent a script tag within an HTML fi
le that references a | |
524 * Dart script. | |
525 * | |
526 * @param nodeStart the token marking the beginning of the tag | |
527 * @param tag the name of the tag | |
528 * @param attributes the attributes in the tag | |
529 * @param attributeEnd the token terminating the region where attributes can b
e | |
530 * @param tagNodes the children of the tag | |
531 * @param contentEnd the token that starts the closing tag | |
532 * @param closingTag the name of the tag that occurs in the closing tag | |
533 * @param nodeEnd the last token in the tag | |
534 */ | |
535 HtmlScriptTagNode(Token nodeStart, Token tag, | |
536 List<XmlAttributeNode> attributes, Token attributeEnd, | |
537 List<XmlTagNode> tagNodes, Token contentEnd, Token closingTag, | |
538 Token nodeEnd) | |
539 : super(nodeStart, tag, attributes, attributeEnd, tagNodes, contentEnd, | |
540 closingTag, nodeEnd); | |
541 | |
542 /** | |
543 * Return the AST structure representing the Dart code within this tag, or `nu
ll` if this | |
544 * tag references an external script. | |
545 * | |
546 * @return the AST structure representing the Dart code within this tag | |
547 */ | |
548 CompilationUnit get script => _script; | |
549 | |
550 /** | |
551 * Set the AST structure representing the Dart code within this tag to the giv
en compilation unit. | |
552 * | |
553 * @param unit the AST structure representing the Dart code within this tag | |
554 */ | |
555 void set script(CompilationUnit unit) { | |
556 _script = unit; | |
557 } | |
558 | |
559 @override | |
560 accept(XmlVisitor visitor) => visitor.visitHtmlScriptTagNode(this); | |
561 } | |
562 | |
563 /** | |
564 * Instances of the class `HtmlUnit` represent the contents of an HTML file. | |
565 */ | |
566 @deprecated | |
567 class HtmlUnit extends XmlNode { | |
568 /** | |
569 * The first token in the token stream that was parsed to form this HTML unit. | |
570 */ | |
571 final Token beginToken; | |
572 | |
573 /** | |
574 * The last token in the token stream that was parsed to form this compilation
unit. This token | |
575 * should always have a type of [TokenType.EOF]. | |
576 */ | |
577 final Token endToken; | |
578 | |
579 /** | |
580 * The tag nodes contained in the receiver (not `null`, contains no `null`s). | |
581 */ | |
582 List<XmlTagNode> _tagNodes; | |
583 | |
584 /** | |
585 * Construct a new instance representing the content of an HTML file. | |
586 * | |
587 * @param beginToken the first token in the file (not `null`) | |
588 * @param tagNodes child tag nodes of the receiver (not `null`, contains no `n
ull`s) | |
589 * @param endToken the last token in the token stream which should be of type | |
590 * [TokenType.EOF] | |
591 */ | |
592 HtmlUnit(this.beginToken, List<XmlTagNode> tagNodes, this.endToken) { | |
593 this._tagNodes = becomeParentOfAll(tagNodes); | |
594 } | |
595 | |
596 /** | |
597 * Return the element associated with this HTML unit. | |
598 * | |
599 * @return the element or `null` if the receiver is not resolved | |
600 */ | |
601 @override | |
602 HtmlElement get element => super.element as HtmlElement; | |
603 | |
604 @override | |
605 void set element(Element element) { | |
606 if (element != null && element is! HtmlElement) { | |
607 throw new IllegalArgumentException( | |
608 "HtmlElement expected, but ${element.runtimeType} given"); | |
609 } | |
610 super.element = element; | |
611 } | |
612 | |
613 /** | |
614 * Answer the tag nodes contained in the receiver. Callers should not manipula
te the returned list | |
615 * to edit the AST structure. | |
616 * | |
617 * @return the children (not `null`, contains no `null`s) | |
618 */ | |
619 List<XmlTagNode> get tagNodes => _tagNodes; | |
620 | |
621 @override | |
622 accept(XmlVisitor visitor) => visitor.visitHtmlUnit(this); | |
623 | |
624 @override | |
625 void visitChildren(XmlVisitor visitor) { | |
626 for (XmlTagNode node in _tagNodes) { | |
627 node.accept(visitor); | |
628 } | |
629 } | |
630 } | |
631 | |
632 /** | |
633 * Instances of the class `RecursiveXmlVisitor` implement an XML visitor that wi
ll recursively | |
634 * visit all of the nodes in an XML structure. For example, using an instance of
this class to visit | |
635 * a [XmlTagNode] will also cause all of the contained [XmlAttributeNode]s and | |
636 * [XmlTagNode]s to be visited. | |
637 * | |
638 * Subclasses that override a visit method must either invoke the overridden vis
it method or must | |
639 * explicitly ask the visited node to visit its children. Failure to do so will
cause the children | |
640 * of the visited node to not be visited. | |
641 */ | |
642 class RecursiveXmlVisitor<R> implements XmlVisitor<R> { | |
643 @override | |
644 R visitHtmlScriptTagNode(HtmlScriptTagNode node) { | |
645 node.visitChildren(this); | |
646 return null; | |
647 } | |
648 | |
649 @override | |
650 R visitHtmlUnit(HtmlUnit node) { | |
651 node.visitChildren(this); | |
652 return null; | |
653 } | |
654 | |
655 @override | |
656 R visitXmlAttributeNode(XmlAttributeNode node) { | |
657 node.visitChildren(this); | |
658 return null; | |
659 } | |
660 | |
661 @override | |
662 R visitXmlTagNode(XmlTagNode node) { | |
663 node.visitChildren(this); | |
664 return null; | |
665 } | |
666 } | |
667 | |
668 /** | |
669 * Instances of the class `SimpleXmlVisitor` implement an AST visitor that will
do nothing | |
670 * when visiting an AST node. It is intended to be a superclass for classes that
use the visitor | |
671 * pattern primarily as a dispatch mechanism (and hence don't need to recursivel
y visit a whole | |
672 * structure) and that only need to visit a small number of node types. | |
673 */ | |
674 class SimpleXmlVisitor<R> implements XmlVisitor<R> { | |
675 @override | |
676 R visitHtmlScriptTagNode(HtmlScriptTagNode node) => null; | |
677 | |
678 @override | |
679 R visitHtmlUnit(HtmlUnit htmlUnit) => null; | |
680 | |
681 @override | |
682 R visitXmlAttributeNode(XmlAttributeNode xmlAttributeNode) => null; | |
683 | |
684 @override | |
685 R visitXmlTagNode(XmlTagNode xmlTagNode) => null; | |
686 } | |
687 | |
688 /** | |
689 * Instances of the class `StringScanner` implement a scanner that reads from a
string. The | |
690 * scanning logic is in the superclass. | |
691 */ | |
692 class StringScanner extends AbstractScanner { | |
693 /** | |
694 * The string from which characters will be read. | |
695 */ | |
696 final String _string; | |
697 | |
698 /** | |
699 * The number of characters in the string. | |
700 */ | |
701 int _stringLength = 0; | |
702 | |
703 /** | |
704 * The index, relative to the string, of the last character that was read. | |
705 */ | |
706 int _charOffset = 0; | |
707 | |
708 /** | |
709 * Initialize a newly created scanner to scan the characters in the given stri
ng. | |
710 * | |
711 * @param source the source being scanned | |
712 * @param string the string from which characters will be read | |
713 */ | |
714 StringScanner(Source source, this._string) : super(source) { | |
715 this._stringLength = _string.length; | |
716 this._charOffset = -1; | |
717 } | |
718 | |
719 @override | |
720 int get offset => _charOffset; | |
721 | |
722 void set offset(int offset) { | |
723 _charOffset = offset; | |
724 } | |
725 | |
726 @override | |
727 int advance() { | |
728 if (++_charOffset < _stringLength) { | |
729 return _string.codeUnitAt(_charOffset); | |
730 } | |
731 _charOffset = _stringLength; | |
732 return -1; | |
733 } | |
734 | |
735 @override | |
736 String getString(int start, int endDelta) => | |
737 _string.substring(start, _charOffset + 1 + endDelta).toString(); | |
738 | |
739 @override | |
740 int peek() { | |
741 if (_charOffset + 1 < _stringLength) { | |
742 return _string.codeUnitAt(_charOffset + 1); | |
743 } | |
744 return -1; | |
745 } | |
746 } | |
747 | |
748 /** | |
749 * Instances of the class `Token` represent a token that was scanned from the in
put. Each | |
750 * token knows which token follows it, acting as the head of a linked list of to
kens. | |
751 */ | |
752 class Token { | |
753 /** | |
754 * The offset from the beginning of the file to the first character in the tok
en. | |
755 */ | |
756 final int offset; | |
757 | |
758 /** | |
759 * The previous token in the token stream. | |
760 */ | |
761 Token previous; | |
762 | |
763 /** | |
764 * The next token in the token stream. | |
765 */ | |
766 Token _next; | |
767 | |
768 /** | |
769 * The type of the token. | |
770 */ | |
771 final TokenType type; | |
772 | |
773 /** | |
774 * The lexeme represented by this token. | |
775 */ | |
776 String _value; | |
777 | |
778 /** | |
779 * Initialize a newly created token. | |
780 * | |
781 * @param type the token type (not `null`) | |
782 * @param offset the offset from the beginning of the file to the first charac
ter in the token | |
783 */ | |
784 Token.con1(TokenType type, int offset) : this.con2(type, offset, type.lexeme); | |
785 | |
786 /** | |
787 * Initialize a newly created token. | |
788 * | |
789 * @param type the token type (not `null`) | |
790 * @param offset the offset from the beginning of the file to the first charac
ter in the token | |
791 * @param value the lexeme represented by this token (not `null`) | |
792 */ | |
793 Token.con2(this.type, this.offset, String value) { | |
794 this._value = StringUtilities.intern(value); | |
795 } | |
796 | |
797 /** | |
798 * Return the offset from the beginning of the file to the character after las
t character of the | |
799 * token. | |
800 * | |
801 * @return the offset from the beginning of the file to the first character af
ter last character | |
802 * of the token | |
803 */ | |
804 int get end => offset + length; | |
805 | |
806 /** | |
807 * Return `true` if this token is a synthetic token. A synthetic token is a to
ken that was | |
808 * introduced by the parser in order to recover from an error in the code. Syn
thetic tokens always | |
809 * have a length of zero (`0`). | |
810 * | |
811 * @return `true` if this token is a synthetic token | |
812 */ | |
813 bool get isSynthetic => length == 0; | |
814 | |
815 /** | |
816 * Return the number of characters in the node's source range. | |
817 * | |
818 * @return the number of characters in the node's source range | |
819 */ | |
820 int get length => lexeme.length; | |
821 | |
822 /** | |
823 * Return the lexeme that represents this token. | |
824 * | |
825 * @return the lexeme (not `null`) | |
826 */ | |
827 String get lexeme => _value; | |
828 | |
829 /** | |
830 * Return the next token in the token stream. | |
831 * | |
832 * @return the next token in the token stream | |
833 */ | |
834 Token get next => _next; | |
835 | |
836 /** | |
837 * Set the next token in the token stream to the given token. This has the sid
e-effect of setting | |
838 * this token to be the previous token for the given token. | |
839 * | |
840 * @param token the next token in the token stream | |
841 * @return the token that was passed in | |
842 */ | |
843 Token setNext(Token token) { | |
844 _next = token; | |
845 token.previous = this; | |
846 return token; | |
847 } | |
848 | |
849 @override | |
850 String toString() => lexeme; | |
851 } | |
852 | |
853 /** | |
854 * The enumeration `TokenType` defines the types of tokens that can be returned
by the | |
855 * scanner. | |
856 */ | |
857 class TokenType extends Enum<TokenType> { | |
858 /** | |
859 * The type of the token that marks the end of the input. | |
860 */ | |
861 static const TokenType EOF = const TokenType_EOF('EOF', 0, ""); | |
862 | |
863 static const TokenType EQ = const TokenType('EQ', 1, "="); | |
864 | |
865 static const TokenType GT = const TokenType('GT', 2, ">"); | |
866 | |
867 static const TokenType LT_SLASH = const TokenType('LT_SLASH', 3, "</"); | |
868 | |
869 static const TokenType LT = const TokenType('LT', 4, "<"); | |
870 | |
871 static const TokenType SLASH_GT = const TokenType('SLASH_GT', 5, "/>"); | |
872 | |
873 static const TokenType COMMENT = const TokenType('COMMENT', 6, null); | |
874 | |
875 static const TokenType DECLARATION = const TokenType('DECLARATION', 7, null); | |
876 | |
877 static const TokenType DIRECTIVE = const TokenType('DIRECTIVE', 8, null); | |
878 | |
879 static const TokenType STRING = const TokenType('STRING', 9, null); | |
880 | |
881 static const TokenType TAG = const TokenType('TAG', 10, null); | |
882 | |
883 static const TokenType TEXT = const TokenType('TEXT', 11, null); | |
884 | |
885 static const List<TokenType> values = const [ | |
886 EOF, | |
887 EQ, | |
888 GT, | |
889 LT_SLASH, | |
890 LT, | |
891 SLASH_GT, | |
892 COMMENT, | |
893 DECLARATION, | |
894 DIRECTIVE, | |
895 STRING, | |
896 TAG, | |
897 TEXT | |
898 ]; | |
899 | |
900 /** | |
901 * The lexeme that defines this type of token, or `null` if there is more than
one possible | |
902 * lexeme for this type of token. | |
903 */ | |
904 final String lexeme; | |
905 | |
906 const TokenType(String name, int ordinal, this.lexeme) : super(name, ordinal); | |
907 } | |
908 | |
909 class TokenType_EOF extends TokenType { | |
910 const TokenType_EOF(String name, int ordinal, String arg0) | |
911 : super(name, ordinal, arg0); | |
912 | |
913 @override | |
914 String toString() => "-eof-"; | |
915 } | |
916 | |
917 /** | |
918 * Instances of the class `ToSourceVisitor` write a source representation of a v
isited XML | |
919 * node (and all of it's children) to a writer. | |
920 */ | |
921 class ToSourceVisitor implements XmlVisitor<Object> { | |
922 /** | |
923 * The writer to which the source is to be written. | |
924 */ | |
925 final PrintWriter _writer; | |
926 | |
927 /** | |
928 * Initialize a newly created visitor to write source code representing the vi
sited nodes to the | |
929 * given writer. | |
930 * | |
931 * @param writer the writer to which the source is to be written | |
932 */ | |
933 ToSourceVisitor(this._writer); | |
934 | |
935 @override | |
936 Object visitHtmlScriptTagNode(HtmlScriptTagNode node) => | |
937 visitXmlTagNode(node); | |
938 | |
939 @override | |
940 Object visitHtmlUnit(HtmlUnit node) { | |
941 for (XmlTagNode child in node.tagNodes) { | |
942 _visit(child); | |
943 } | |
944 return null; | |
945 } | |
946 | |
947 @override | |
948 Object visitXmlAttributeNode(XmlAttributeNode node) { | |
949 String name = node.name; | |
950 Token value = node.valueToken; | |
951 if (name.length == 0) { | |
952 _writer.print("__"); | |
953 } else { | |
954 _writer.print(name); | |
955 } | |
956 _writer.print("="); | |
957 if (value == null) { | |
958 _writer.print("__"); | |
959 } else { | |
960 _writer.print(value.lexeme); | |
961 } | |
962 return null; | |
963 } | |
964 | |
965 @override | |
966 Object visitXmlTagNode(XmlTagNode node) { | |
967 _writer.print("<"); | |
968 String tagName = node.tag; | |
969 _writer.print(tagName); | |
970 for (XmlAttributeNode attribute in node.attributes) { | |
971 _writer.print(" "); | |
972 _visit(attribute); | |
973 } | |
974 _writer.print(node.attributeEnd.lexeme); | |
975 if (node.closingTag != null) { | |
976 for (XmlTagNode child in node.tagNodes) { | |
977 _visit(child); | |
978 } | |
979 _writer.print("</"); | |
980 _writer.print(tagName); | |
981 _writer.print(">"); | |
982 } | |
983 return null; | |
984 } | |
985 | |
986 /** | |
987 * Safely visit the given node. | |
988 * | |
989 * @param node the node to be visited | |
990 */ | |
991 void _visit(XmlNode node) { | |
992 if (node != null) { | |
993 node.accept(this); | |
994 } | |
995 } | |
996 } | |
997 | |
998 /** | |
999 * Instances of `XmlAttributeNode` represent name/value pairs owned by an [XmlTa
gNode]. | |
1000 */ | |
1001 class XmlAttributeNode extends XmlNode { | |
1002 /** | |
1003 * An empty list of XML attribute nodes. | |
1004 */ | |
1005 static const List<XmlAttributeNode> EMPTY_LIST = const <XmlAttributeNode>[]; | |
1006 | |
1007 final Token _name; | |
1008 | |
1009 final Token equals; | |
1010 | |
1011 final Token _value; | |
1012 | |
1013 List<XmlExpression> expressions = XmlExpression.EMPTY_ARRAY; | |
1014 | |
1015 /** | |
1016 * Construct a new instance representing an XML attribute. | |
1017 * | |
1018 * @param name the name token (not `null`). This may be a zero length token if
the attribute | |
1019 * is badly formed. | |
1020 * @param equals the equals sign or `null` if none | |
1021 * @param value the value token (not `null`) | |
1022 */ | |
1023 XmlAttributeNode(this._name, this.equals, this._value); | |
1024 | |
1025 @override | |
1026 Token get beginToken => _name; | |
1027 | |
1028 @override | |
1029 Token get endToken => _value; | |
1030 | |
1031 /** | |
1032 * Answer the attribute name. This may be a zero length string if the attribut
e is badly formed. | |
1033 * | |
1034 * @return the name (not `null`) | |
1035 */ | |
1036 String get name => _name.lexeme; | |
1037 | |
1038 /** | |
1039 * Answer the attribute name token. This may be a zero length token if the att
ribute is badly | |
1040 * formed. | |
1041 * | |
1042 * @return the name token (not `null`) | |
1043 */ | |
1044 Token get nameToken => _name; | |
1045 | |
1046 /** | |
1047 * Answer the lexeme for the value token without the leading and trailing quot
es. | |
1048 * | |
1049 * @return the text or `null` if the value is not specified | |
1050 */ | |
1051 String get text { | |
1052 if (_value == null) { | |
1053 return null; | |
1054 } | |
1055 //TODO (danrubel): replace HTML character encodings with the actual | |
1056 // characters | |
1057 String text = _value.lexeme; | |
1058 int len = text.length; | |
1059 if (len > 0) { | |
1060 if (text.codeUnitAt(0) == 0x22) { | |
1061 if (len > 1 && text.codeUnitAt(len - 1) == 0x22) { | |
1062 return text.substring(1, len - 1); | |
1063 } else { | |
1064 return text.substring(1); | |
1065 } | |
1066 } else if (text.codeUnitAt(0) == 0x27) { | |
1067 if (len > 1 && text.codeUnitAt(len - 1) == 0x27) { | |
1068 return text.substring(1, len - 1); | |
1069 } else { | |
1070 return text.substring(1); | |
1071 } | |
1072 } | |
1073 } | |
1074 return text; | |
1075 } | |
1076 | |
1077 /** | |
1078 * Answer the offset of the value after the leading quote. | |
1079 * | |
1080 * @return the offset of the value, or `-1` if the value is not specified | |
1081 */ | |
1082 int get textOffset { | |
1083 if (_value == null) { | |
1084 return -1; | |
1085 } | |
1086 String text = _value.lexeme; | |
1087 if (StringUtilities.startsWithChar(text, 0x22) || | |
1088 StringUtilities.startsWithChar(text, 0x27)) { | |
1089 return _value.offset + 1; | |
1090 } | |
1091 return _value.offset; | |
1092 } | |
1093 | |
1094 /** | |
1095 * Answer the attribute value token. A properly formed value will start and en
d with matching | |
1096 * quote characters, but the value returned may not be properly formed. | |
1097 * | |
1098 * @return the value token or `null` if this represents a badly formed attribu
te | |
1099 */ | |
1100 Token get valueToken => _value; | |
1101 | |
1102 @override | |
1103 accept(XmlVisitor visitor) => visitor.visitXmlAttributeNode(this); | |
1104 | |
1105 @override | |
1106 void visitChildren(XmlVisitor visitor) { | |
1107 // no children to visit | |
1108 } | |
1109 } | |
1110 | |
1111 /** | |
1112 * Instances of the class `XmlExpression` represent an abstract expression embed
ded into | |
1113 * [XmlNode]. | |
1114 */ | |
1115 abstract class XmlExpression { | |
1116 /** | |
1117 * An empty list of expressions. | |
1118 */ | |
1119 static const List<XmlExpression> EMPTY_ARRAY = const <XmlExpression>[]; | |
1120 | |
1121 /** | |
1122 * Return the offset of the character immediately following the last character
of this | |
1123 * expression's source range. This is equivalent to `getOffset() + getLength()
`. | |
1124 * | |
1125 * @return the offset of the character just past the expression's source range | |
1126 */ | |
1127 int get end; | |
1128 | |
1129 /** | |
1130 * Return the number of characters in the expression's source range. | |
1131 */ | |
1132 int get length; | |
1133 | |
1134 /** | |
1135 * Return the offset of the first character in the expression's source range. | |
1136 */ | |
1137 int get offset; | |
1138 | |
1139 /** | |
1140 * Check if the given offset belongs to the expression's source range. | |
1141 */ | |
1142 bool contains(int offset) => this.offset <= offset && offset < end; | |
1143 | |
1144 /** | |
1145 * Return the [Reference] at the given offset. | |
1146 * | |
1147 * @param offset the offset from the beginning of the file | |
1148 * @return the [Reference] at the given offset, maybe `null` | |
1149 */ | |
1150 XmlExpression_Reference getReference(int offset); | |
1151 } | |
1152 | |
1153 /** | |
1154 * The reference to the [Element]. | |
1155 */ | |
1156 class XmlExpression_Reference { | |
1157 Element element; | |
1158 | |
1159 int offset = 0; | |
1160 | |
1161 int length = 0; | |
1162 | |
1163 XmlExpression_Reference(Element element, int offset, int length) { | |
1164 this.element = element; | |
1165 this.offset = offset; | |
1166 this.length = length; | |
1167 } | |
1168 } | |
1169 | |
1170 /** | |
1171 * The abstract class `XmlNode` defines behavior common to all XML/HTML nodes. | |
1172 */ | |
1173 abstract class XmlNode { | |
1174 /** | |
1175 * The parent of the node, or `null` if the node is the root of an AST structu
re. | |
1176 */ | |
1177 XmlNode _parent; | |
1178 | |
1179 /** | |
1180 * The element associated with this node or `null` if the receiver is not reso
lved. | |
1181 */ | |
1182 Element _element; | |
1183 | |
1184 /** | |
1185 * Return the first token included in this node's source range. | |
1186 * | |
1187 * @return the first token or `null` if none | |
1188 */ | |
1189 Token get beginToken; | |
1190 | |
1191 /** | |
1192 * Return the element associated with this node. | |
1193 * | |
1194 * @return the element or `null` if the receiver is not resolved | |
1195 */ | |
1196 Element get element => _element; | |
1197 | |
1198 /** | |
1199 * Set the element associated with this node. | |
1200 * | |
1201 * @param element the element | |
1202 */ | |
1203 void set element(Element element) { | |
1204 this._element = element; | |
1205 } | |
1206 | |
1207 /** | |
1208 * Return the offset of the character immediately following the last character
of this node's | |
1209 * source range. This is equivalent to `node.getOffset() + node.getLength()`.
For an html | |
1210 * unit this will be equal to the length of the unit's source. | |
1211 * | |
1212 * @return the offset of the character just past the node's source range | |
1213 */ | |
1214 int get end => offset + length; | |
1215 | |
1216 /** | |
1217 * Return the last token included in this node's source range. | |
1218 * | |
1219 * @return the last token or `null` if none | |
1220 */ | |
1221 Token get endToken; | |
1222 | |
1223 /** | |
1224 * Return the number of characters in the node's source range. | |
1225 * | |
1226 * @return the number of characters in the node's source range | |
1227 */ | |
1228 int get length { | |
1229 Token beginToken = this.beginToken; | |
1230 Token endToken = this.endToken; | |
1231 if (beginToken == null || endToken == null) { | |
1232 return -1; | |
1233 } | |
1234 return endToken.offset + endToken.length - beginToken.offset; | |
1235 } | |
1236 | |
1237 /** | |
1238 * Return the offset from the beginning of the file to the first character in
the node's source | |
1239 * range. | |
1240 * | |
1241 * @return the offset from the beginning of the file to the first character in
the node's source | |
1242 * range | |
1243 */ | |
1244 int get offset { | |
1245 Token beginToken = this.beginToken; | |
1246 if (beginToken == null) { | |
1247 return -1; | |
1248 } | |
1249 return this.beginToken.offset; | |
1250 } | |
1251 | |
1252 /** | |
1253 * Return this node's parent node, or `null` if this node is the root of an AS
T structure. | |
1254 * | |
1255 * Note that the relationship between an AST node and its parent node may chan
ge over the lifetime | |
1256 * of a node. | |
1257 * | |
1258 * @return the parent of this node, or `null` if none | |
1259 */ | |
1260 XmlNode get parent => _parent; | |
1261 | |
1262 /** | |
1263 * Set the parent of this node to the given node. | |
1264 * | |
1265 * @param newParent the node that is to be made the parent of this node | |
1266 */ | |
1267 void set parent(XmlNode newParent) { | |
1268 XmlNode current = newParent; | |
1269 while (current != null) { | |
1270 if (identical(current, this)) { | |
1271 AnalysisEngine.instance.logger.logError( | |
1272 "Circular structure while setting an XML node's parent", | |
1273 new CaughtException( | |
1274 new ArgumentError(_buildRecursiveStructureMessage(newParent)), | |
1275 null)); | |
1276 return; | |
1277 } | |
1278 current = current.parent; | |
1279 } | |
1280 _parent = newParent; | |
1281 } | |
1282 | |
1283 /** | |
1284 * Use the given visitor to visit this node. | |
1285 * | |
1286 * @param visitor the visitor that will visit this node | |
1287 * @return the value returned by the visitor as a result of visiting this node | |
1288 */ | |
1289 accept(XmlVisitor visitor); | |
1290 | |
1291 /** | |
1292 * Make this node the parent of the given child node. | |
1293 * | |
1294 * @param child the node that will become a child of this node | |
1295 * @return the node that was made a child of this node | |
1296 */ | |
1297 XmlNode becomeParentOf(XmlNode child) { | |
1298 if (child != null) { | |
1299 XmlNode node = child; | |
1300 node.parent = this; | |
1301 } | |
1302 return child; | |
1303 } | |
1304 | |
1305 /** | |
1306 * Make this node the parent of the given child nodes. | |
1307 * | |
1308 * @param children the nodes that will become the children of this node | |
1309 * @param ifEmpty the (empty) nodes to return if "children" is empty | |
1310 * @return the nodes that were made children of this node | |
1311 */ | |
1312 List becomeParentOfAll(List children, {List ifEmpty}) { | |
1313 if (children == null || children.isEmpty) { | |
1314 if (ifEmpty != null) { | |
1315 return ifEmpty; | |
1316 } | |
1317 } | |
1318 if (children != null) { | |
1319 children.forEach((XmlNode node) { | |
1320 node.parent = this; | |
1321 }); | |
1322 } | |
1323 return children; | |
1324 } | |
1325 | |
1326 @override | |
1327 String toString() { | |
1328 PrintStringWriter writer = new PrintStringWriter(); | |
1329 accept(new ToSourceVisitor(writer)); | |
1330 return writer.toString(); | |
1331 } | |
1332 | |
1333 /** | |
1334 * Use the given visitor to visit all of the children of this node. The childr
en will be visited | |
1335 * in source order. | |
1336 * | |
1337 * @param visitor the visitor that will be used to visit the children of this
node | |
1338 */ | |
1339 void visitChildren(XmlVisitor visitor); | |
1340 | |
1341 /** | |
1342 * This method exists for debugging purposes only. | |
1343 */ | |
1344 void _appendIdentifier(StringBuffer buffer, XmlNode node) { | |
1345 if (node is XmlTagNode) { | |
1346 buffer.write(node.tag); | |
1347 } else if (node is XmlAttributeNode) { | |
1348 buffer.write(node.name); | |
1349 } else { | |
1350 buffer.write("htmlUnit"); | |
1351 } | |
1352 } | |
1353 | |
1354 /** | |
1355 * This method exists for debugging purposes only. | |
1356 */ | |
1357 String _buildRecursiveStructureMessage(XmlNode newParent) { | |
1358 StringBuffer buffer = new StringBuffer(); | |
1359 buffer.write("Attempt to create recursive structure: "); | |
1360 XmlNode current = newParent; | |
1361 while (current != null) { | |
1362 if (!identical(current, newParent)) { | |
1363 buffer.write(" -> "); | |
1364 } | |
1365 if (identical(current, this)) { | |
1366 buffer.writeCharCode(0x2A); | |
1367 _appendIdentifier(buffer, current); | |
1368 buffer.writeCharCode(0x2A); | |
1369 } else { | |
1370 _appendIdentifier(buffer, current); | |
1371 } | |
1372 current = current.parent; | |
1373 } | |
1374 return buffer.toString(); | |
1375 } | |
1376 } | |
1377 | |
1378 /** | |
1379 * Instances of the class `XmlParser` are used to parse tokens into a AST struct
ure comprised | |
1380 * of [XmlNode]s. | |
1381 */ | |
1382 class XmlParser { | |
1383 /** | |
1384 * The source being parsed. | |
1385 */ | |
1386 final Source source; | |
1387 | |
1388 /** | |
1389 * The next token to be parsed. | |
1390 */ | |
1391 Token _currentToken; | |
1392 | |
1393 /** | |
1394 * Construct a parser for the specified source. | |
1395 * | |
1396 * @param source the source being parsed | |
1397 */ | |
1398 XmlParser(this.source); | |
1399 | |
1400 /** | |
1401 * Answer the current token. | |
1402 * | |
1403 * @return the current token | |
1404 */ | |
1405 Token get currentToken => _currentToken; | |
1406 | |
1407 /** | |
1408 * Create a node representing an attribute. | |
1409 * | |
1410 * @param name the name of the attribute | |
1411 * @param equals the equals sign, or `null` if there is no value | |
1412 * @param value the value of the attribute | |
1413 * @return the node that was created | |
1414 */ | |
1415 XmlAttributeNode createAttributeNode(Token name, Token equals, Token value) => | |
1416 new XmlAttributeNode(name, equals, value); | |
1417 | |
1418 /** | |
1419 * Create a node representing a tag. | |
1420 * | |
1421 * @param nodeStart the token marking the beginning of the tag | |
1422 * @param tag the name of the tag | |
1423 * @param attributes the attributes in the tag | |
1424 * @param attributeEnd the token terminating the region where attributes can b
e | |
1425 * @param tagNodes the children of the tag | |
1426 * @param contentEnd the token that starts the closing tag | |
1427 * @param closingTag the name of the tag that occurs in the closing tag | |
1428 * @param nodeEnd the last token in the tag | |
1429 * @return the node that was created | |
1430 */ | |
1431 XmlTagNode createTagNode(Token nodeStart, Token tag, | |
1432 List<XmlAttributeNode> attributes, Token attributeEnd, | |
1433 List<XmlTagNode> tagNodes, Token contentEnd, Token closingTag, | |
1434 Token nodeEnd) => new XmlTagNode(nodeStart, tag, attributes, attributeEnd, | |
1435 tagNodes, contentEnd, closingTag, nodeEnd); | |
1436 | |
1437 /** | |
1438 * Answer `true` if the specified tag is self closing and thus should never ha
ve content or | |
1439 * child tag nodes. | |
1440 * | |
1441 * @param tag the tag (not `null`) | |
1442 * @return `true` if self closing | |
1443 */ | |
1444 bool isSelfClosing(Token tag) => false; | |
1445 | |
1446 /** | |
1447 * Parse the entire token stream and in the process, advance the current token
to the end of the | |
1448 * token stream. | |
1449 * | |
1450 * @return the list of tag nodes found (not `null`, contains no `null`) | |
1451 */ | |
1452 List<XmlTagNode> parseTopTagNodes(Token firstToken) { | |
1453 _currentToken = firstToken; | |
1454 List<XmlTagNode> tagNodes = new List<XmlTagNode>(); | |
1455 TokenType type = _currentToken.type; | |
1456 while (type != TokenType.EOF) { | |
1457 if (type == TokenType.LT) { | |
1458 tagNodes.add(_parseTagNode()); | |
1459 } else if (type == TokenType.DECLARATION || | |
1460 type == TokenType.DIRECTIVE || | |
1461 type == TokenType.COMMENT) { | |
1462 // ignored tokens | |
1463 _currentToken = _currentToken.next; | |
1464 } else { | |
1465 _reportUnexpectedToken(); | |
1466 _currentToken = _currentToken.next; | |
1467 } | |
1468 type = _currentToken.type; | |
1469 } | |
1470 return tagNodes; | |
1471 } | |
1472 | |
1473 /** | |
1474 * Insert a synthetic token of the specified type before the current token | |
1475 * | |
1476 * @param type the type of token to be inserted (not `null`) | |
1477 * @return the synthetic token that was inserted (not `null`) | |
1478 */ | |
1479 Token _insertSyntheticToken(TokenType type) { | |
1480 Token token = new Token.con2(type, _currentToken.offset, ""); | |
1481 _currentToken.previous.setNext(token); | |
1482 token.setNext(_currentToken); | |
1483 return token; | |
1484 } | |
1485 | |
1486 /** | |
1487 * Parse the token stream for an attribute. This method advances the current t
oken over the | |
1488 * attribute, but should not be called if the [currentToken] is not [TokenType
.TAG]. | |
1489 * | |
1490 * @return the attribute (not `null`) | |
1491 */ | |
1492 XmlAttributeNode _parseAttribute() { | |
1493 // Assume the current token is a tag | |
1494 Token name = _currentToken; | |
1495 _currentToken = _currentToken.next; | |
1496 // Equals sign | |
1497 Token equals; | |
1498 if (_currentToken.type == TokenType.EQ) { | |
1499 equals = _currentToken; | |
1500 _currentToken = _currentToken.next; | |
1501 } else { | |
1502 _reportUnexpectedToken(); | |
1503 equals = _insertSyntheticToken(TokenType.EQ); | |
1504 } | |
1505 // String value | |
1506 Token value; | |
1507 if (_currentToken.type == TokenType.STRING) { | |
1508 value = _currentToken; | |
1509 _currentToken = _currentToken.next; | |
1510 } else { | |
1511 _reportUnexpectedToken(); | |
1512 value = _insertSyntheticToken(TokenType.STRING); | |
1513 } | |
1514 return createAttributeNode(name, equals, value); | |
1515 } | |
1516 | |
1517 /** | |
1518 * Parse the stream for a sequence of attributes. This method advances the cur
rent token to the | |
1519 * next [TokenType.GT], [TokenType.SLASH_GT], or [TokenType.EOF]. | |
1520 * | |
1521 * @return a collection of zero or more attributes (not `null`, contains no `n
ull`s) | |
1522 */ | |
1523 List<XmlAttributeNode> _parseAttributes() { | |
1524 TokenType type = _currentToken.type; | |
1525 if (type == TokenType.GT || | |
1526 type == TokenType.SLASH_GT || | |
1527 type == TokenType.EOF) { | |
1528 return XmlTagNode.NO_ATTRIBUTES; | |
1529 } | |
1530 List<XmlAttributeNode> attributes = new List<XmlAttributeNode>(); | |
1531 while (type != TokenType.GT && | |
1532 type != TokenType.SLASH_GT && | |
1533 type != TokenType.EOF) { | |
1534 if (type == TokenType.TAG) { | |
1535 attributes.add(_parseAttribute()); | |
1536 } else { | |
1537 _reportUnexpectedToken(); | |
1538 _currentToken = _currentToken.next; | |
1539 } | |
1540 type = _currentToken.type; | |
1541 } | |
1542 return attributes; | |
1543 } | |
1544 | |
1545 /** | |
1546 * Parse the stream for a sequence of tag nodes existing within a parent tag n
ode. This method | |
1547 * advances the current token to the next [TokenType.LT_SLASH] or [TokenType.E
OF]. | |
1548 * | |
1549 * @return a list of nodes (not `null`, contains no `null`s) | |
1550 */ | |
1551 List<XmlTagNode> _parseChildTagNodes() { | |
1552 TokenType type = _currentToken.type; | |
1553 if (type == TokenType.LT_SLASH || type == TokenType.EOF) { | |
1554 return XmlTagNode.NO_TAG_NODES; | |
1555 } | |
1556 List<XmlTagNode> nodes = new List<XmlTagNode>(); | |
1557 while (type != TokenType.LT_SLASH && type != TokenType.EOF) { | |
1558 if (type == TokenType.LT) { | |
1559 nodes.add(_parseTagNode()); | |
1560 } else if (type == TokenType.COMMENT) { | |
1561 // ignored token | |
1562 _currentToken = _currentToken.next; | |
1563 } else { | |
1564 _reportUnexpectedToken(); | |
1565 _currentToken = _currentToken.next; | |
1566 } | |
1567 type = _currentToken.type; | |
1568 } | |
1569 return nodes; | |
1570 } | |
1571 | |
1572 /** | |
1573 * Parse the token stream for the next tag node. This method advances current
token over the | |
1574 * parsed tag node, but should only be called if the current token is [TokenTy
pe.LT] | |
1575 * | |
1576 * @return the tag node or `null` if none found | |
1577 */ | |
1578 XmlTagNode _parseTagNode() { | |
1579 // Assume that the current node is a tag node start TokenType.LT | |
1580 Token nodeStart = _currentToken; | |
1581 _currentToken = _currentToken.next; | |
1582 // Get the tag or create a synthetic tag and report an error | |
1583 Token tag; | |
1584 if (_currentToken.type == TokenType.TAG) { | |
1585 tag = _currentToken; | |
1586 _currentToken = _currentToken.next; | |
1587 } else { | |
1588 _reportUnexpectedToken(); | |
1589 tag = _insertSyntheticToken(TokenType.TAG); | |
1590 } | |
1591 // Parse the attributes | |
1592 List<XmlAttributeNode> attributes = _parseAttributes(); | |
1593 // Token ending attribute list | |
1594 Token attributeEnd; | |
1595 if (_currentToken.type == TokenType.GT || | |
1596 _currentToken.type == TokenType.SLASH_GT) { | |
1597 attributeEnd = _currentToken; | |
1598 _currentToken = _currentToken.next; | |
1599 } else { | |
1600 _reportUnexpectedToken(); | |
1601 attributeEnd = _insertSyntheticToken(TokenType.SLASH_GT); | |
1602 } | |
1603 // If the node has no children, then return the node | |
1604 if (attributeEnd.type == TokenType.SLASH_GT || isSelfClosing(tag)) { | |
1605 return createTagNode(nodeStart, tag, attributes, attributeEnd, | |
1606 XmlTagNode.NO_TAG_NODES, _currentToken, null, attributeEnd); | |
1607 } | |
1608 // Parse the child tag nodes | |
1609 List<XmlTagNode> tagNodes = _parseChildTagNodes(); | |
1610 // Token ending child tag nodes | |
1611 Token contentEnd; | |
1612 if (_currentToken.type == TokenType.LT_SLASH) { | |
1613 contentEnd = _currentToken; | |
1614 _currentToken = _currentToken.next; | |
1615 } else { | |
1616 // TODO (danrubel): handle self closing HTML elements by inserting | |
1617 // synthetic tokens but not reporting an error | |
1618 _reportUnexpectedToken(); | |
1619 contentEnd = _insertSyntheticToken(TokenType.LT_SLASH); | |
1620 } | |
1621 // Closing tag | |
1622 Token closingTag; | |
1623 if (_currentToken.type == TokenType.TAG) { | |
1624 closingTag = _currentToken; | |
1625 _currentToken = _currentToken.next; | |
1626 } else { | |
1627 _reportUnexpectedToken(); | |
1628 closingTag = _insertSyntheticToken(TokenType.TAG); | |
1629 } | |
1630 // Token ending node | |
1631 Token nodeEnd; | |
1632 if (_currentToken.type == TokenType.GT) { | |
1633 nodeEnd = _currentToken; | |
1634 _currentToken = _currentToken.next; | |
1635 } else { | |
1636 _reportUnexpectedToken(); | |
1637 nodeEnd = _insertSyntheticToken(TokenType.GT); | |
1638 } | |
1639 return createTagNode(nodeStart, tag, attributes, attributeEnd, tagNodes, | |
1640 contentEnd, closingTag, nodeEnd); | |
1641 } | |
1642 | |
1643 /** | |
1644 * Report the current token as unexpected | |
1645 */ | |
1646 void _reportUnexpectedToken() { | |
1647 // TODO (danrubel): report unexpected token | |
1648 } | |
1649 } | |
1650 | |
1651 /** | |
1652 * Instances of `XmlTagNode` represent XML or HTML elements such as `` and | |
1653 * `<body foo="bar"> ... </body>`. | |
1654 */ | |
1655 class XmlTagNode extends XmlNode { | |
1656 /** | |
1657 * Constant representing empty list of attributes. | |
1658 */ | |
1659 static List<XmlAttributeNode> NO_ATTRIBUTES = | |
1660 new UnmodifiableListView(new List<XmlAttributeNode>()); | |
1661 | |
1662 /** | |
1663 * Constant representing empty list of tag nodes. | |
1664 */ | |
1665 static List<XmlTagNode> NO_TAG_NODES = | |
1666 new UnmodifiableListView(new List<XmlTagNode>()); | |
1667 | |
1668 /** | |
1669 * The starting [TokenType.LT] token (not `null`). | |
1670 */ | |
1671 final Token nodeStart; | |
1672 | |
1673 /** | |
1674 * The [TokenType.TAG] token after the starting '<' (not `null`). | |
1675 */ | |
1676 final Token _tag; | |
1677 | |
1678 /** | |
1679 * The attributes contained by the receiver (not `null`, contains no `null`s). | |
1680 */ | |
1681 List<XmlAttributeNode> _attributes; | |
1682 | |
1683 /** | |
1684 * The [TokenType.GT] or [TokenType.SLASH_GT] token after the attributes (not | |
1685 * `null`). The token may be the same token as [nodeEnd] if there are no child | |
1686 * [tagNodes]. | |
1687 */ | |
1688 final Token attributeEnd; | |
1689 | |
1690 /** | |
1691 * The tag nodes contained in the receiver (not `null`, contains no `null`s). | |
1692 */ | |
1693 List<XmlTagNode> _tagNodes; | |
1694 | |
1695 /** | |
1696 * The token (not `null`) after the content, which may be | |
1697 * * (1) [TokenType.LT_SLASH] for nodes with open and close tags, or | |
1698 * * (2) the [TokenType.LT] nodeStart of the next sibling node if this node is
self | |
1699 * closing or the attributeEnd is [TokenType.SLASH_GT], or | |
1700 * * (3) [TokenType.EOF] if the node does not have a closing tag and is the la
st node in | |
1701 * the stream [TokenType.LT_SLASH] token after the content, or `null` if there
is no | |
1702 * content and the attributes ended with [TokenType.SLASH_GT]. | |
1703 */ | |
1704 final Token contentEnd; | |
1705 | |
1706 /** | |
1707 * The closing [TokenType.TAG] after the child elements or `null` if there is
no | |
1708 * content and the attributes ended with [TokenType.SLASH_GT] | |
1709 */ | |
1710 final Token closingTag; | |
1711 | |
1712 /** | |
1713 * The ending [TokenType.GT] or [TokenType.SLASH_GT] token (not `null`). | |
1714 */ | |
1715 final Token nodeEnd; | |
1716 | |
1717 /** | |
1718 * The expressions that are embedded in the tag's content. | |
1719 */ | |
1720 List<XmlExpression> expressions = XmlExpression.EMPTY_ARRAY; | |
1721 | |
1722 /** | |
1723 * Construct a new instance representing an XML or HTML element | |
1724 * | |
1725 * @param nodeStart the starting [TokenType.LT] token (not `null`) | |
1726 * @param tag the [TokenType.TAG] token after the starting '<' (not `null`)
. | |
1727 * @param attributes the attributes associated with this element or [NO_ATTRIB
UTES] (not | |
1728 * `null`, contains no `null`s) | |
1729 * @param attributeEnd The [TokenType.GT] or [TokenType.SLASH_GT] token after
the | |
1730 * attributes (not `null`). The token may be the same token as [nodeE
nd] if | |
1731 * there are no child [tagNodes]. | |
1732 * @param tagNodes child tag nodes of the receiver or [NO_TAG_NODES] (not `nul
l`, | |
1733 * contains no `null`s) | |
1734 * @param contentEnd the token (not `null`) after the content, which may be | |
1735 * * (1) [TokenType.LT_SLASH] for nodes with open and close tags, or | |
1736 * * (2) the [TokenType.LT] nodeStart of the next sibling node if thi
s node is | |
1737 * self closing or the attributeEnd is [TokenType.SLASH_GT], or | |
1738 * * (3) [TokenType.EOF] if the node does not have a closing tag and
is the last | |
1739 * node in the stream [TokenType.LT_SLASH] token after the content, o
r `null` | |
1740 * if there is no content and the attributes ended with [TokenType.SL
ASH_GT]. | |
1741 * @param closingTag the closing [TokenType.TAG] after the child elements or `
null` if | |
1742 * there is no content and the attributes ended with [TokenType.SLASH
_GT] | |
1743 * @param nodeEnd the ending [TokenType.GT] or [TokenType.SLASH_GT] token (not | |
1744 * `null`) | |
1745 */ | |
1746 XmlTagNode(this.nodeStart, this._tag, List<XmlAttributeNode> attributes, | |
1747 this.attributeEnd, List<XmlTagNode> tagNodes, this.contentEnd, | |
1748 this.closingTag, this.nodeEnd) { | |
1749 this._attributes = becomeParentOfAll(attributes, ifEmpty: NO_ATTRIBUTES); | |
1750 this._tagNodes = becomeParentOfAll(tagNodes, ifEmpty: NO_TAG_NODES); | |
1751 } | |
1752 | |
1753 /** | |
1754 * Answer the receiver's attributes. Callers should not manipulate the returne
d list to edit the | |
1755 * AST structure. | |
1756 * | |
1757 * @return the attributes (not `null`, contains no `null`s) | |
1758 */ | |
1759 List<XmlAttributeNode> get attributes => _attributes; | |
1760 | |
1761 @override | |
1762 Token get beginToken => nodeStart; | |
1763 | |
1764 /** | |
1765 * Return a string representing the content contained in the receiver. This | |
1766 * includes the textual representation of any child tag nodes ([getTagNodes]). | |
1767 * Whitespace between '<', '</', and '>', '/>' is discarded, but all | |
1768 * other whitespace is preserved. | |
1769 */ | |
1770 String get content { | |
1771 Token token = attributeEnd.next; | |
1772 if (identical(token, contentEnd)) { | |
1773 return ""; | |
1774 } | |
1775 // TODO(danrubel) Handle CDATA and replace HTML character encodings with | |
1776 // the actual characters. | |
1777 String content = token.lexeme; | |
1778 token = token.next; | |
1779 if (identical(token, contentEnd)) { | |
1780 return content; | |
1781 } | |
1782 StringBuffer buffer = new StringBuffer(); | |
1783 buffer.write(content); | |
1784 while (!identical(token, contentEnd)) { | |
1785 buffer.write(token.lexeme); | |
1786 token = token.next; | |
1787 } | |
1788 return buffer.toString(); | |
1789 } | |
1790 | |
1791 @override | |
1792 Token get endToken { | |
1793 if (nodeEnd != null) { | |
1794 return nodeEnd; | |
1795 } | |
1796 if (closingTag != null) { | |
1797 return closingTag; | |
1798 } | |
1799 if (contentEnd != null) { | |
1800 return contentEnd; | |
1801 } | |
1802 if (!_tagNodes.isEmpty) { | |
1803 return _tagNodes[_tagNodes.length - 1].endToken; | |
1804 } | |
1805 if (attributeEnd != null) { | |
1806 return attributeEnd; | |
1807 } | |
1808 if (!_attributes.isEmpty) { | |
1809 return _attributes[_attributes.length - 1].endToken; | |
1810 } | |
1811 return _tag; | |
1812 } | |
1813 | |
1814 /** | |
1815 * Answer the tag name after the starting '<'. | |
1816 * | |
1817 * @return the tag name (not `null`) | |
1818 */ | |
1819 String get tag => _tag.lexeme; | |
1820 | |
1821 /** | |
1822 * Answer the tag nodes contained in the receiver. Callers should not manipula
te the returned list | |
1823 * to edit the AST structure. | |
1824 * | |
1825 * @return the children (not `null`, contains no `null`s) | |
1826 */ | |
1827 List<XmlTagNode> get tagNodes => _tagNodes; | |
1828 | |
1829 /** | |
1830 * Answer the [TokenType.TAG] token after the starting '<'. | |
1831 * | |
1832 * @return the token (not `null`) | |
1833 */ | |
1834 Token get tagToken => _tag; | |
1835 | |
1836 @override | |
1837 accept(XmlVisitor visitor) => visitor.visitXmlTagNode(this); | |
1838 | |
1839 /** | |
1840 * Answer the attribute with the specified name. | |
1841 * | |
1842 * @param name the attribute name | |
1843 * @return the attribute or `null` if no matching attribute is found | |
1844 */ | |
1845 XmlAttributeNode getAttribute(String name) { | |
1846 for (XmlAttributeNode attribute in _attributes) { | |
1847 if (attribute.name == name) { | |
1848 return attribute; | |
1849 } | |
1850 } | |
1851 return null; | |
1852 } | |
1853 | |
1854 /** | |
1855 * Find the attribute with the given name (see [getAttribute] and answer the l
exeme | |
1856 * for the attribute's value token without the leading and trailing quotes (se
e | |
1857 * [XmlAttributeNode.getText]). | |
1858 * | |
1859 * @param name the attribute name | |
1860 * @return the attribute text or `null` if no matching attribute is found | |
1861 */ | |
1862 String getAttributeText(String name) { | |
1863 XmlAttributeNode attribute = getAttribute(name); | |
1864 return attribute != null ? attribute.text : null; | |
1865 } | |
1866 | |
1867 @override | |
1868 void visitChildren(XmlVisitor visitor) { | |
1869 for (XmlAttributeNode node in _attributes) { | |
1870 node.accept(visitor); | |
1871 } | |
1872 for (XmlTagNode node in _tagNodes) { | |
1873 node.accept(visitor); | |
1874 } | |
1875 } | |
1876 } | |
1877 | |
1878 /** | |
1879 * The interface `XmlVisitor` defines the behavior of objects that can be used t
o visit an | |
1880 * [XmlNode] structure. | |
1881 */ | |
1882 abstract class XmlVisitor<R> { | |
1883 R visitHtmlScriptTagNode(HtmlScriptTagNode node); | |
1884 | |
1885 R visitHtmlUnit(HtmlUnit htmlUnit); | |
1886 | |
1887 R visitXmlAttributeNode(XmlAttributeNode xmlAttributeNode); | |
1888 | |
1889 R visitXmlTagNode(XmlTagNode xmlTagNode); | |
1890 } | |
OLD | NEW |