1 /*
2 * Copyright (c) 2010, 2015, Oracle and/or its affiliates. All rights reserved.
3 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
4 *
5 * This code is free software; you can redistribute it and/or modify it
6 * under the terms of the GNU General Public License version 2 only, as
7 * published by the Free Software Foundation. Oracle designates this
8 * particular file as subject to the "Classpath" exception as provided
9 * by Oracle in the LICENSE file that accompanied this code.
10 *
11 * This code is distributed in the hope that it will be useful, but WITHOUT
12 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
13 * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
14 * version 2 for more details (a copy is included in the LICENSE file that
15 * accompanied this code).
16 *
17 * You should have received a copy of the GNU General Public License version
18 * 2 along with this work; if not, write to the Free Software Foundation,
19 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
20 *
21 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
22 * or visit www.oracle.com if you need additional information or have any
23 * questions.
24 */
25
26 package jdk.nashorn.internal.parser;
27
28 import java.util.ArrayList;
29 import java.util.List;
30 import jdk.nashorn.internal.codegen.ObjectClassGenerator;
31 import jdk.nashorn.internal.objects.Global;
32 import jdk.nashorn.internal.runtime.ECMAErrors;
33 import jdk.nashorn.internal.runtime.ErrorManager;
34 import jdk.nashorn.internal.runtime.JSErrorType;
35 import jdk.nashorn.internal.runtime.JSType;
36 import jdk.nashorn.internal.runtime.ParserException;
37 import jdk.nashorn.internal.runtime.Property;
38 import jdk.nashorn.internal.runtime.PropertyMap;
39 import jdk.nashorn.internal.runtime.ScriptObject;
40 import jdk.nashorn.internal.runtime.Source;
41 import jdk.nashorn.internal.runtime.SpillProperty;
42 import jdk.nashorn.internal.runtime.arrays.ArrayData;
43 import jdk.nashorn.internal.runtime.arrays.ArrayIndex;
44 import jdk.nashorn.internal.scripts.JO;
45
46 import static jdk.nashorn.internal.parser.TokenType.STRING;
47
48 /**
49 * Parses JSON text and returns the corresponding IR node. This is derived from the objectLiteral production of the main parser.
50 *
51 * See: 15.12.1.2 The JSON Syntactic Grammar
52 */
53 public class JSONParser {
54
55 final private String source;
56 final private Global global;
57 final int length;
58 int pos = 0;
59
60 private static PropertyMap EMPTY_MAP = PropertyMap.newMap();
61
62 private static final int EOF = -1;
63
64 private static final String TRUE = "true";
65 private static final String FALSE = "false";
66 private static final String NULL = "null";
67
68 private static final int STATE_EMPTY = 0;
69 private static final int STATE_ELEMENT_PARSED = 1;
70 private static final int STATE_COMMA_PARSED = 2;
71
72 /**
73 * Constructor
74 * @param source the source
75 * @param global the global object
76 */
77 public JSONParser(final String source, final Global global ) {
78 this.source = source;
79 this.global = global;
80 this.length = source.length();
81 }
82
83 /**
84 * Implementation of the Quote(value) operation as defined in the ECMA script spec
85 * It wraps a String value in double quotes and escapes characters within in
86 *
87 * @param value string to quote
88 *
89 * @return quoted and escaped string
90 */
91 public static String quote(final String value) {
92
93 final StringBuilder product = new StringBuilder();
94
95 product.append("\"");
96
97 for (final char ch : value.toCharArray()) {
98 // TODO: should use a table?
99 switch (ch) {
100 case '\\':
101 product.append("\\\\");
102 break;
103 case '"':
104 product.append("\\\"");
105 break;
106 case '\b':
107 product.append("\\b");
108 break;
109 case '\f':
110 product.append("\\f");
111 break;
112 case '\n':
113 product.append("\\n");
114 break;
115 case '\r':
116 product.append("\\r");
117 break;
118 case '\t':
119 product.append("\\t");
120 break;
121 default:
122 if (ch < ' ') {
123 product.append(Lexer.unicodeEscape(ch));
124 break;
125 }
126
127 product.append(ch);
128 break;
129 }
130 }
131
132 product.append("\"");
133
134 return product.toString();
135 }
136
137 /**
138 * Public parse method. Parse a string into a JSON object.
139 *
140 * @return the parsed JSON Object
141 */
142 public Object parse() {
143 final Object value = parseLiteral();
144 skipWhiteSpace();
145 if (pos < length) {
146 throw expectedError(pos, "eof", toString(peek()));
147 }
148 return value;
149 }
150
151 private Object parseLiteral() {
152 skipWhiteSpace();
153
154 final int c = peek();
155 if (c == EOF) {
156 throw expectedError(pos, "json literal", "eof");
157 }
158 switch (c) {
159 case '{':
160 return parseObject();
161 case '[':
162 return parseArray();
163 case '"':
164 return parseString();
165 case 'f':
166 return parseKeyword(FALSE, Boolean.FALSE);
167 case 't':
168 return parseKeyword(TRUE, Boolean.TRUE);
169 case 'n':
170 return parseKeyword(NULL, null);
171 default:
172 if (isDigit(c) || c == '-') {
173 return parseNumber();
174 } else if (c == '.') {
175 throw numberError(pos);
176 } else {
177 throw expectedError(pos, "json literal", toString(c));
178 }
179 }
180 }
181
182 private Object parseObject() {
183 PropertyMap propertyMap = EMPTY_MAP;
184 ArrayData arrayData = ArrayData.EMPTY_ARRAY;
185 final ArrayList<Object> values = new ArrayList<>();
186 int state = STATE_EMPTY;
187
188 assert peek() == '{';
189 pos++;
190
191 while (pos < length) {
192 skipWhiteSpace();
193 final int c = peek();
194
195 switch (c) {
196 case '"':
197 if (state == STATE_ELEMENT_PARSED) {
198 throw expectedError(pos - 1, ", or }", toString(c));
199 }
200 final String id = parseString();
201 expectColon();
202 final Object value = parseLiteral();
203 final int index = ArrayIndex.getArrayIndex(id);
204 if (ArrayIndex.isValidArrayIndex(index)) {
205 arrayData = addArrayElement(arrayData, index, value);
206 } else {
207 propertyMap = addObjectProperty(propertyMap, values, id, value);
208 }
209 state = STATE_ELEMENT_PARSED;
210 break;
211 case ',':
212 if (state != STATE_ELEMENT_PARSED) {
213 throw error(AbstractParser.message("trailing.comma.in.json"), pos);
214 }
215 state = STATE_COMMA_PARSED;
216 pos++;
217 break;
218 case '}':
219 if (state == STATE_COMMA_PARSED) {
220 throw error(AbstractParser.message("trailing.comma.in.json"), pos);
221 }
222 pos++;
223 return createObject(propertyMap, values, arrayData);
224 default:
225 throw expectedError(pos, ", or }", toString(c));
226 }
227 }
228 throw expectedError(pos, ", or }", "eof");
229 }
230
231 private static ArrayData addArrayElement(final ArrayData arrayData, final int index, final Object value) {
232 final long oldLength = arrayData.length();
233 final long longIndex = ArrayIndex.toLongIndex(index);
234 ArrayData newArrayData = arrayData;
235 if (longIndex > oldLength) {
236 if (arrayData.canDelete(oldLength, longIndex - 1, false)) {
237 newArrayData = newArrayData.delete(oldLength, longIndex - 1);
238 }
239 }
240 return newArrayData.ensure(longIndex).set(index, value, false);
241 }
242
243 private static PropertyMap addObjectProperty(final PropertyMap propertyMap, final List<Object> values,
244 final String id, final Object value) {
245 final Property oldProperty = propertyMap.findProperty(id);
246 final Property newProperty;
247 final PropertyMap newMap;
248 final Class<?> type = ObjectClassGenerator.OBJECT_FIELDS_ONLY ? Object.class : getType(value);
249
250 if (oldProperty != null) {
251 values.set(oldProperty.getSlot(), value);
252 newProperty = new SpillProperty(id, 0, oldProperty.getSlot());
253 newProperty.setType(type);
254 newMap = propertyMap.replaceProperty(oldProperty, newProperty);;
255 } else {
256 values.add(value);
257 newProperty = new SpillProperty(id, 0, propertyMap.size());
258 newProperty.setType(type);
259 newMap = propertyMap.addProperty(newProperty);
260 }
261
262 return newMap;
263 }
264
265 private Object createObject(final PropertyMap propertyMap, final List<Object> values, final ArrayData arrayData) {
266 final long[] primitiveSpill = new long[values.size()];
267 final Object[] objectSpill = new Object[values.size()];
268
269 for (final Property property : propertyMap.getProperties()) {
270 if (property.getType() == Object.class) {
271 objectSpill[property.getSlot()] = values.get(property.getSlot());
272 } else {
273 primitiveSpill[property.getSlot()] = ObjectClassGenerator.pack((Number) values.get(property.getSlot()));
274 }
275 }
276
277 final ScriptObject object = new JO(propertyMap, primitiveSpill, objectSpill);
278 object.setInitialProto(global.getObjectPrototype());
279 object.setArray(arrayData);
280 return object;
281 }
282
283 private static Class<?> getType(final Object value) {
284 if (value instanceof Integer) {
285 return int.class;
286 } else if (value instanceof Long) {
287 return long.class;
288 } else if (value instanceof Double) {
289 return double.class;
290 } else {
291 return Object.class;
292 }
293 }
294
295 private void expectColon() {
296 skipWhiteSpace();
297 final int n = next();
298 if (n != ':') {
299 throw expectedError(pos - 1, ":", toString(n));
300 }
301 }
302
303 private Object parseArray() {
304 ArrayData arrayData = ArrayData.EMPTY_ARRAY;
305 int state = STATE_EMPTY;
306
307 assert peek() == '[';
308 pos++;
309
310 while (pos < length) {
311 skipWhiteSpace();
312 final int c = peek();
313
314 switch (c) {
315 case ',':
316 if (state != STATE_ELEMENT_PARSED) {
317 throw error(AbstractParser.message("trailing.comma.in.json"), pos);
318 }
319 state = STATE_COMMA_PARSED;
320 pos++;
321 break;
322 case ']':
323 if (state == STATE_COMMA_PARSED) {
324 throw error(AbstractParser.message("trailing.comma.in.json"), pos);
325 }
326 pos++;
327 return global.wrapAsObject(arrayData);
328 default:
329 if (state == STATE_ELEMENT_PARSED) {
330 throw expectedError(pos, ", or ]", toString(c));
331 }
332 final long index = arrayData.length();
333 arrayData = arrayData.ensure(index).set((int) index, parseLiteral(), true);
334 state = STATE_ELEMENT_PARSED;
335 break;
336 }
337 }
338
339 throw expectedError(pos, ", or ]", "eof");
340 }
341
342 private String parseString() {
343 // String buffer is only instantiated if string contains escape sequences.
344 int start = ++pos;
345 StringBuilder sb = null;
346
347 while (pos < length) {
348 final int c = next();
349 if (c <= 0x1f) {
350 // Characters < 0x1f are not allowed in JSON strings.
351 throw syntaxError(pos, "String contains control character");
352
353 } else if (c == '\\') {
354 if (sb == null) {
355 sb = new StringBuilder(pos - start + 16);
356 }
357 sb.append(source, start, pos - 1);
358 sb.append(parseEscapeSequence());
359 start = pos;
360
361 } else if (c == '"') {
362 if (sb != null) {
363 sb.append(source, start, pos - 1);
364 return sb.toString();
365 }
366 return source.substring(start, pos - 1);
367 }
368 }
369
370 throw error(Lexer.message("missing.close.quote"), pos, length);
371 }
372
373 private char parseEscapeSequence() {
374 final int c = next();
375 switch (c) {
376 case '"':
377 return '"';
378 case '\\':
379 return '\\';
380 case '/':
381 return '/';
382 case 'b':
383 return '\b';
384 case 'f':
385 return '\f';
386 case 'n':
387 return '\n';
388 case 'r':
389 return '\r';
390 case 't':
391 return '\t';
392 case 'u':
393 return parseUnicodeEscape();
394 default:
395 throw error(Lexer.message("invalid.escape.char"), pos - 1, length);
396 }
397 }
398
399 private char parseUnicodeEscape() {
400 return (char) (parseHexDigit() << 12 | parseHexDigit() << 8 | parseHexDigit() << 4 | parseHexDigit());
401 }
402
403 private int parseHexDigit() {
404 final int c = next();
405 if (c >= '0' && c <= '9') {
406 return c - '0';
407 } else if (c >= 'A' && c <= 'F') {
408 return c + 10 - 'A';
409 } else if (c >= 'a' && c <= 'f') {
410 return c + 10 - 'a';
411 }
412 throw error(Lexer.message("invalid.hex"), pos - 1, length);
413 }
414
415 private boolean isDigit(final int c) {
416 return c >= '0' && c <= '9';
417 }
418
419 private void skipDigits() {
420 while (pos < length) {
421 final int c = peek();
422 if (!isDigit(c)) {
423 break;
424 }
425 pos++;
426 }
427 }
428
429 private Number parseNumber() {
430 final int start = pos;
431 int c = next();
432
433 if (c == '-') {
434 c = next();
435 }
436 if (!isDigit(c)) {
437 throw numberError(start);
438 }
439 // no more digits allowed after 0
440 if (c != '0') {
441 skipDigits();
442 }
443
444 // fraction
445 if (peek() == '.') {
446 pos++;
447 if (!isDigit(next())) {
448 throw numberError(pos - 1);
449 }
450 skipDigits();
451 }
452
453 // exponent
454 c = peek();
455 if (c == 'e' || c == 'E') {
456 pos++;
457 c = next();
458 if (c == '-' || c == '+') {
459 c = next();
460 }
461 if (!isDigit(c)) {
462 throw numberError(pos - 1);
463 }
464 skipDigits();
465 }
466
467 final double d = Double.parseDouble(source.substring(start, pos));
468 if (JSType.isRepresentableAsInt(d)) {
469 return (int) d;
470 } else if (JSType.isRepresentableAsLong(d)) {
471 return (long) d;
472 }
473 return d;
474 }
475
476 private Object parseKeyword(final String keyword, final Object value) {
477 if (!source.regionMatches(pos, keyword, 0, keyword.length())) {
478 throw expectedError(pos, "json literal", "ident");
479 }
480 pos += keyword.length();
481 return value;
482 }
483
484 private int peek() {
485 if (pos >= length) {
486 return -1;
487 }
488 return source.charAt(pos);
489 }
490
491 private int next() {
492 final int next = peek();
493 pos++;
494 return next;
495 }
496
497 private void skipWhiteSpace() {
498 while (pos < length) {
499 switch (peek()) {
500 case '\t':
501 case '\r':
502 case '\n':
503 case ' ':
504 pos++;
505 break;
506 default:
507 return;
508 }
509 }
510 }
511
512 private static String toString(final int c) {
513 return c == EOF ? "eof" : String.valueOf((char) c);
514 }
515
516 ParserException error(final String message, final int start, final int length) throws ParserException {
517 final long token = Token.toDesc(STRING, start, length);
518 final int pos = Token.descPosition(token);
519 final Source src = Source.sourceFor("<json>", source);
520 final int lineNum = src.getLine(pos);
521 final int columnNum = src.getColumn(pos);
522 final String formatted = ErrorManager.format(message, src, lineNum, columnNum, token);
523 return new ParserException(JSErrorType.SYNTAX_ERROR, formatted, src, lineNum, columnNum, token);
524 }
525
526 private ParserException error(final String message, final int start) {
527 return error(message, start, length);
528 }
529
530 private ParserException numberError(final int start) {
531 return error(Lexer.message("json.invalid.number"), start);
532 }
533
534 private ParserException expectedError(final int start, final String expected, final String found) {
535 return error(AbstractParser.message("expected", expected, found), start);
536 }
537
538 private ParserException syntaxError(final int start, final String reason) {
539 final String message = ECMAErrors.getMessage("syntax.error.invalid.json", reason);
540 return error(message, start);
541 }
542 }