1 | var Tokenizer = require('./tokenizer');
|
---|
2 | var TAB = 9;
|
---|
3 | var N = 10;
|
---|
4 | var F = 12;
|
---|
5 | var R = 13;
|
---|
6 | var SPACE = 32;
|
---|
7 | var EXCLAMATIONMARK = 33; // !
|
---|
8 | var NUMBERSIGN = 35; // #
|
---|
9 | var AMPERSAND = 38; // &
|
---|
10 | var APOSTROPHE = 39; // '
|
---|
11 | var LEFTPARENTHESIS = 40; // (
|
---|
12 | var RIGHTPARENTHESIS = 41; // )
|
---|
13 | var ASTERISK = 42; // *
|
---|
14 | var PLUSSIGN = 43; // +
|
---|
15 | var COMMA = 44; // ,
|
---|
16 | var HYPERMINUS = 45; // -
|
---|
17 | var LESSTHANSIGN = 60; // <
|
---|
18 | var GREATERTHANSIGN = 62; // >
|
---|
19 | var QUESTIONMARK = 63; // ?
|
---|
20 | var COMMERCIALAT = 64; // @
|
---|
21 | var LEFTSQUAREBRACKET = 91; // [
|
---|
22 | var RIGHTSQUAREBRACKET = 93; // ]
|
---|
23 | var LEFTCURLYBRACKET = 123; // {
|
---|
24 | var VERTICALLINE = 124; // |
|
---|
25 | var RIGHTCURLYBRACKET = 125; // }
|
---|
26 | var INFINITY = 8734; // ∞
|
---|
27 | var NAME_CHAR = createCharMap(function(ch) {
|
---|
28 | return /[a-zA-Z0-9\-]/.test(ch);
|
---|
29 | });
|
---|
30 | var COMBINATOR_PRECEDENCE = {
|
---|
31 | ' ': 1,
|
---|
32 | '&&': 2,
|
---|
33 | '||': 3,
|
---|
34 | '|': 4
|
---|
35 | };
|
---|
36 |
|
---|
37 | function createCharMap(fn) {
|
---|
38 | var array = typeof Uint32Array === 'function' ? new Uint32Array(128) : new Array(128);
|
---|
39 | for (var i = 0; i < 128; i++) {
|
---|
40 | array[i] = fn(String.fromCharCode(i)) ? 1 : 0;
|
---|
41 | }
|
---|
42 | return array;
|
---|
43 | }
|
---|
44 |
|
---|
45 | function scanSpaces(tokenizer) {
|
---|
46 | return tokenizer.substringToPos(
|
---|
47 | tokenizer.findWsEnd(tokenizer.pos)
|
---|
48 | );
|
---|
49 | }
|
---|
50 |
|
---|
51 | function scanWord(tokenizer) {
|
---|
52 | var end = tokenizer.pos;
|
---|
53 |
|
---|
54 | for (; end < tokenizer.str.length; end++) {
|
---|
55 | var code = tokenizer.str.charCodeAt(end);
|
---|
56 | if (code >= 128 || NAME_CHAR[code] === 0) {
|
---|
57 | break;
|
---|
58 | }
|
---|
59 | }
|
---|
60 |
|
---|
61 | if (tokenizer.pos === end) {
|
---|
62 | tokenizer.error('Expect a keyword');
|
---|
63 | }
|
---|
64 |
|
---|
65 | return tokenizer.substringToPos(end);
|
---|
66 | }
|
---|
67 |
|
---|
68 | function scanNumber(tokenizer) {
|
---|
69 | var end = tokenizer.pos;
|
---|
70 |
|
---|
71 | for (; end < tokenizer.str.length; end++) {
|
---|
72 | var code = tokenizer.str.charCodeAt(end);
|
---|
73 | if (code < 48 || code > 57) {
|
---|
74 | break;
|
---|
75 | }
|
---|
76 | }
|
---|
77 |
|
---|
78 | if (tokenizer.pos === end) {
|
---|
79 | tokenizer.error('Expect a number');
|
---|
80 | }
|
---|
81 |
|
---|
82 | return tokenizer.substringToPos(end);
|
---|
83 | }
|
---|
84 |
|
---|
85 | function scanString(tokenizer) {
|
---|
86 | var end = tokenizer.str.indexOf('\'', tokenizer.pos + 1);
|
---|
87 |
|
---|
88 | if (end === -1) {
|
---|
89 | tokenizer.pos = tokenizer.str.length;
|
---|
90 | tokenizer.error('Expect an apostrophe');
|
---|
91 | }
|
---|
92 |
|
---|
93 | return tokenizer.substringToPos(end + 1);
|
---|
94 | }
|
---|
95 |
|
---|
96 | function readMultiplierRange(tokenizer) {
|
---|
97 | var min = null;
|
---|
98 | var max = null;
|
---|
99 |
|
---|
100 | tokenizer.eat(LEFTCURLYBRACKET);
|
---|
101 |
|
---|
102 | min = scanNumber(tokenizer);
|
---|
103 |
|
---|
104 | if (tokenizer.charCode() === COMMA) {
|
---|
105 | tokenizer.pos++;
|
---|
106 | if (tokenizer.charCode() !== RIGHTCURLYBRACKET) {
|
---|
107 | max = scanNumber(tokenizer);
|
---|
108 | }
|
---|
109 | } else {
|
---|
110 | max = min;
|
---|
111 | }
|
---|
112 |
|
---|
113 | tokenizer.eat(RIGHTCURLYBRACKET);
|
---|
114 |
|
---|
115 | return {
|
---|
116 | min: Number(min),
|
---|
117 | max: max ? Number(max) : 0
|
---|
118 | };
|
---|
119 | }
|
---|
120 |
|
---|
121 | function readMultiplier(tokenizer) {
|
---|
122 | var range = null;
|
---|
123 | var comma = false;
|
---|
124 |
|
---|
125 | switch (tokenizer.charCode()) {
|
---|
126 | case ASTERISK:
|
---|
127 | tokenizer.pos++;
|
---|
128 |
|
---|
129 | range = {
|
---|
130 | min: 0,
|
---|
131 | max: 0
|
---|
132 | };
|
---|
133 |
|
---|
134 | break;
|
---|
135 |
|
---|
136 | case PLUSSIGN:
|
---|
137 | tokenizer.pos++;
|
---|
138 |
|
---|
139 | range = {
|
---|
140 | min: 1,
|
---|
141 | max: 0
|
---|
142 | };
|
---|
143 |
|
---|
144 | break;
|
---|
145 |
|
---|
146 | case QUESTIONMARK:
|
---|
147 | tokenizer.pos++;
|
---|
148 |
|
---|
149 | range = {
|
---|
150 | min: 0,
|
---|
151 | max: 1
|
---|
152 | };
|
---|
153 |
|
---|
154 | break;
|
---|
155 |
|
---|
156 | case NUMBERSIGN:
|
---|
157 | tokenizer.pos++;
|
---|
158 |
|
---|
159 | comma = true;
|
---|
160 |
|
---|
161 | if (tokenizer.charCode() === LEFTCURLYBRACKET) {
|
---|
162 | range = readMultiplierRange(tokenizer);
|
---|
163 | } else {
|
---|
164 | range = {
|
---|
165 | min: 1,
|
---|
166 | max: 0
|
---|
167 | };
|
---|
168 | }
|
---|
169 |
|
---|
170 | break;
|
---|
171 |
|
---|
172 | case LEFTCURLYBRACKET:
|
---|
173 | range = readMultiplierRange(tokenizer);
|
---|
174 | break;
|
---|
175 |
|
---|
176 | default:
|
---|
177 | return null;
|
---|
178 | }
|
---|
179 |
|
---|
180 | return {
|
---|
181 | type: 'Multiplier',
|
---|
182 | comma: comma,
|
---|
183 | min: range.min,
|
---|
184 | max: range.max,
|
---|
185 | term: null
|
---|
186 | };
|
---|
187 | }
|
---|
188 |
|
---|
189 | function maybeMultiplied(tokenizer, node) {
|
---|
190 | var multiplier = readMultiplier(tokenizer);
|
---|
191 |
|
---|
192 | if (multiplier !== null) {
|
---|
193 | multiplier.term = node;
|
---|
194 | return multiplier;
|
---|
195 | }
|
---|
196 |
|
---|
197 | return node;
|
---|
198 | }
|
---|
199 |
|
---|
200 | function maybeToken(tokenizer) {
|
---|
201 | var ch = tokenizer.peek();
|
---|
202 |
|
---|
203 | if (ch === '') {
|
---|
204 | return null;
|
---|
205 | }
|
---|
206 |
|
---|
207 | return {
|
---|
208 | type: 'Token',
|
---|
209 | value: ch
|
---|
210 | };
|
---|
211 | }
|
---|
212 |
|
---|
213 | function readProperty(tokenizer) {
|
---|
214 | var name;
|
---|
215 |
|
---|
216 | tokenizer.eat(LESSTHANSIGN);
|
---|
217 | tokenizer.eat(APOSTROPHE);
|
---|
218 |
|
---|
219 | name = scanWord(tokenizer);
|
---|
220 |
|
---|
221 | tokenizer.eat(APOSTROPHE);
|
---|
222 | tokenizer.eat(GREATERTHANSIGN);
|
---|
223 |
|
---|
224 | return maybeMultiplied(tokenizer, {
|
---|
225 | type: 'Property',
|
---|
226 | name: name
|
---|
227 | });
|
---|
228 | }
|
---|
229 |
|
---|
230 | // https://drafts.csswg.org/css-values-3/#numeric-ranges
|
---|
231 | // 4.1. Range Restrictions and Range Definition Notation
|
---|
232 | //
|
---|
233 | // Range restrictions can be annotated in the numeric type notation using CSS bracketed
|
---|
234 | // range notation—[min,max]—within the angle brackets, after the identifying keyword,
|
---|
235 | // indicating a closed range between (and including) min and max.
|
---|
236 | // For example, <integer [0, 10]> indicates an integer between 0 and 10, inclusive.
|
---|
237 | function readTypeRange(tokenizer) {
|
---|
238 | // use null for Infinity to make AST format JSON serializable/deserializable
|
---|
239 | var min = null; // -Infinity
|
---|
240 | var max = null; // Infinity
|
---|
241 | var sign = 1;
|
---|
242 |
|
---|
243 | tokenizer.eat(LEFTSQUAREBRACKET);
|
---|
244 |
|
---|
245 | if (tokenizer.charCode() === HYPERMINUS) {
|
---|
246 | tokenizer.peek();
|
---|
247 | sign = -1;
|
---|
248 | }
|
---|
249 |
|
---|
250 | if (sign == -1 && tokenizer.charCode() === INFINITY) {
|
---|
251 | tokenizer.peek();
|
---|
252 | } else {
|
---|
253 | min = sign * Number(scanNumber(tokenizer));
|
---|
254 | }
|
---|
255 |
|
---|
256 | scanSpaces(tokenizer);
|
---|
257 | tokenizer.eat(COMMA);
|
---|
258 | scanSpaces(tokenizer);
|
---|
259 |
|
---|
260 | if (tokenizer.charCode() === INFINITY) {
|
---|
261 | tokenizer.peek();
|
---|
262 | } else {
|
---|
263 | sign = 1;
|
---|
264 |
|
---|
265 | if (tokenizer.charCode() === HYPERMINUS) {
|
---|
266 | tokenizer.peek();
|
---|
267 | sign = -1;
|
---|
268 | }
|
---|
269 |
|
---|
270 | max = sign * Number(scanNumber(tokenizer));
|
---|
271 | }
|
---|
272 |
|
---|
273 | tokenizer.eat(RIGHTSQUAREBRACKET);
|
---|
274 |
|
---|
275 | // If no range is indicated, either by using the bracketed range notation
|
---|
276 | // or in the property description, then [−∞,∞] is assumed.
|
---|
277 | if (min === null && max === null) {
|
---|
278 | return null;
|
---|
279 | }
|
---|
280 |
|
---|
281 | return {
|
---|
282 | type: 'Range',
|
---|
283 | min: min,
|
---|
284 | max: max
|
---|
285 | };
|
---|
286 | }
|
---|
287 |
|
---|
288 | function readType(tokenizer) {
|
---|
289 | var name;
|
---|
290 | var opts = null;
|
---|
291 |
|
---|
292 | tokenizer.eat(LESSTHANSIGN);
|
---|
293 | name = scanWord(tokenizer);
|
---|
294 |
|
---|
295 | if (tokenizer.charCode() === LEFTPARENTHESIS &&
|
---|
296 | tokenizer.nextCharCode() === RIGHTPARENTHESIS) {
|
---|
297 | tokenizer.pos += 2;
|
---|
298 | name += '()';
|
---|
299 | }
|
---|
300 |
|
---|
301 | if (tokenizer.charCodeAt(tokenizer.findWsEnd(tokenizer.pos)) === LEFTSQUAREBRACKET) {
|
---|
302 | scanSpaces(tokenizer);
|
---|
303 | opts = readTypeRange(tokenizer);
|
---|
304 | }
|
---|
305 |
|
---|
306 | tokenizer.eat(GREATERTHANSIGN);
|
---|
307 |
|
---|
308 | return maybeMultiplied(tokenizer, {
|
---|
309 | type: 'Type',
|
---|
310 | name: name,
|
---|
311 | opts: opts
|
---|
312 | });
|
---|
313 | }
|
---|
314 |
|
---|
315 | function readKeywordOrFunction(tokenizer) {
|
---|
316 | var name;
|
---|
317 |
|
---|
318 | name = scanWord(tokenizer);
|
---|
319 |
|
---|
320 | if (tokenizer.charCode() === LEFTPARENTHESIS) {
|
---|
321 | tokenizer.pos++;
|
---|
322 |
|
---|
323 | return {
|
---|
324 | type: 'Function',
|
---|
325 | name: name
|
---|
326 | };
|
---|
327 | }
|
---|
328 |
|
---|
329 | return maybeMultiplied(tokenizer, {
|
---|
330 | type: 'Keyword',
|
---|
331 | name: name
|
---|
332 | });
|
---|
333 | }
|
---|
334 |
|
---|
335 | function regroupTerms(terms, combinators) {
|
---|
336 | function createGroup(terms, combinator) {
|
---|
337 | return {
|
---|
338 | type: 'Group',
|
---|
339 | terms: terms,
|
---|
340 | combinator: combinator,
|
---|
341 | disallowEmpty: false,
|
---|
342 | explicit: false
|
---|
343 | };
|
---|
344 | }
|
---|
345 |
|
---|
346 | combinators = Object.keys(combinators).sort(function(a, b) {
|
---|
347 | return COMBINATOR_PRECEDENCE[a] - COMBINATOR_PRECEDENCE[b];
|
---|
348 | });
|
---|
349 |
|
---|
350 | while (combinators.length > 0) {
|
---|
351 | var combinator = combinators.shift();
|
---|
352 | for (var i = 0, subgroupStart = 0; i < terms.length; i++) {
|
---|
353 | var term = terms[i];
|
---|
354 | if (term.type === 'Combinator') {
|
---|
355 | if (term.value === combinator) {
|
---|
356 | if (subgroupStart === -1) {
|
---|
357 | subgroupStart = i - 1;
|
---|
358 | }
|
---|
359 | terms.splice(i, 1);
|
---|
360 | i--;
|
---|
361 | } else {
|
---|
362 | if (subgroupStart !== -1 && i - subgroupStart > 1) {
|
---|
363 | terms.splice(
|
---|
364 | subgroupStart,
|
---|
365 | i - subgroupStart,
|
---|
366 | createGroup(terms.slice(subgroupStart, i), combinator)
|
---|
367 | );
|
---|
368 | i = subgroupStart + 1;
|
---|
369 | }
|
---|
370 | subgroupStart = -1;
|
---|
371 | }
|
---|
372 | }
|
---|
373 | }
|
---|
374 |
|
---|
375 | if (subgroupStart !== -1 && combinators.length) {
|
---|
376 | terms.splice(
|
---|
377 | subgroupStart,
|
---|
378 | i - subgroupStart,
|
---|
379 | createGroup(terms.slice(subgroupStart, i), combinator)
|
---|
380 | );
|
---|
381 | }
|
---|
382 | }
|
---|
383 |
|
---|
384 | return combinator;
|
---|
385 | }
|
---|
386 |
|
---|
387 | function readImplicitGroup(tokenizer) {
|
---|
388 | var terms = [];
|
---|
389 | var combinators = {};
|
---|
390 | var token;
|
---|
391 | var prevToken = null;
|
---|
392 | var prevTokenPos = tokenizer.pos;
|
---|
393 |
|
---|
394 | while (token = peek(tokenizer)) {
|
---|
395 | if (token.type !== 'Spaces') {
|
---|
396 | if (token.type === 'Combinator') {
|
---|
397 | // check for combinator in group beginning and double combinator sequence
|
---|
398 | if (prevToken === null || prevToken.type === 'Combinator') {
|
---|
399 | tokenizer.pos = prevTokenPos;
|
---|
400 | tokenizer.error('Unexpected combinator');
|
---|
401 | }
|
---|
402 |
|
---|
403 | combinators[token.value] = true;
|
---|
404 | } else if (prevToken !== null && prevToken.type !== 'Combinator') {
|
---|
405 | combinators[' '] = true; // a b
|
---|
406 | terms.push({
|
---|
407 | type: 'Combinator',
|
---|
408 | value: ' '
|
---|
409 | });
|
---|
410 | }
|
---|
411 |
|
---|
412 | terms.push(token);
|
---|
413 | prevToken = token;
|
---|
414 | prevTokenPos = tokenizer.pos;
|
---|
415 | }
|
---|
416 | }
|
---|
417 |
|
---|
418 | // check for combinator in group ending
|
---|
419 | if (prevToken !== null && prevToken.type === 'Combinator') {
|
---|
420 | tokenizer.pos -= prevTokenPos;
|
---|
421 | tokenizer.error('Unexpected combinator');
|
---|
422 | }
|
---|
423 |
|
---|
424 | return {
|
---|
425 | type: 'Group',
|
---|
426 | terms: terms,
|
---|
427 | combinator: regroupTerms(terms, combinators) || ' ',
|
---|
428 | disallowEmpty: false,
|
---|
429 | explicit: false
|
---|
430 | };
|
---|
431 | }
|
---|
432 |
|
---|
433 | function readGroup(tokenizer) {
|
---|
434 | var result;
|
---|
435 |
|
---|
436 | tokenizer.eat(LEFTSQUAREBRACKET);
|
---|
437 | result = readImplicitGroup(tokenizer);
|
---|
438 | tokenizer.eat(RIGHTSQUAREBRACKET);
|
---|
439 |
|
---|
440 | result.explicit = true;
|
---|
441 |
|
---|
442 | if (tokenizer.charCode() === EXCLAMATIONMARK) {
|
---|
443 | tokenizer.pos++;
|
---|
444 | result.disallowEmpty = true;
|
---|
445 | }
|
---|
446 |
|
---|
447 | return result;
|
---|
448 | }
|
---|
449 |
|
---|
450 | function peek(tokenizer) {
|
---|
451 | var code = tokenizer.charCode();
|
---|
452 |
|
---|
453 | if (code < 128 && NAME_CHAR[code] === 1) {
|
---|
454 | return readKeywordOrFunction(tokenizer);
|
---|
455 | }
|
---|
456 |
|
---|
457 | switch (code) {
|
---|
458 | case RIGHTSQUAREBRACKET:
|
---|
459 | // don't eat, stop scan a group
|
---|
460 | break;
|
---|
461 |
|
---|
462 | case LEFTSQUAREBRACKET:
|
---|
463 | return maybeMultiplied(tokenizer, readGroup(tokenizer));
|
---|
464 |
|
---|
465 | case LESSTHANSIGN:
|
---|
466 | return tokenizer.nextCharCode() === APOSTROPHE
|
---|
467 | ? readProperty(tokenizer)
|
---|
468 | : readType(tokenizer);
|
---|
469 |
|
---|
470 | case VERTICALLINE:
|
---|
471 | return {
|
---|
472 | type: 'Combinator',
|
---|
473 | value: tokenizer.substringToPos(
|
---|
474 | tokenizer.nextCharCode() === VERTICALLINE
|
---|
475 | ? tokenizer.pos + 2
|
---|
476 | : tokenizer.pos + 1
|
---|
477 | )
|
---|
478 | };
|
---|
479 |
|
---|
480 | case AMPERSAND:
|
---|
481 | tokenizer.pos++;
|
---|
482 | tokenizer.eat(AMPERSAND);
|
---|
483 |
|
---|
484 | return {
|
---|
485 | type: 'Combinator',
|
---|
486 | value: '&&'
|
---|
487 | };
|
---|
488 |
|
---|
489 | case COMMA:
|
---|
490 | tokenizer.pos++;
|
---|
491 | return {
|
---|
492 | type: 'Comma'
|
---|
493 | };
|
---|
494 |
|
---|
495 | case APOSTROPHE:
|
---|
496 | return maybeMultiplied(tokenizer, {
|
---|
497 | type: 'String',
|
---|
498 | value: scanString(tokenizer)
|
---|
499 | });
|
---|
500 |
|
---|
501 | case SPACE:
|
---|
502 | case TAB:
|
---|
503 | case N:
|
---|
504 | case R:
|
---|
505 | case F:
|
---|
506 | return {
|
---|
507 | type: 'Spaces',
|
---|
508 | value: scanSpaces(tokenizer)
|
---|
509 | };
|
---|
510 |
|
---|
511 | case COMMERCIALAT:
|
---|
512 | code = tokenizer.nextCharCode();
|
---|
513 |
|
---|
514 | if (code < 128 && NAME_CHAR[code] === 1) {
|
---|
515 | tokenizer.pos++;
|
---|
516 | return {
|
---|
517 | type: 'AtKeyword',
|
---|
518 | name: scanWord(tokenizer)
|
---|
519 | };
|
---|
520 | }
|
---|
521 |
|
---|
522 | return maybeToken(tokenizer);
|
---|
523 |
|
---|
524 | case ASTERISK:
|
---|
525 | case PLUSSIGN:
|
---|
526 | case QUESTIONMARK:
|
---|
527 | case NUMBERSIGN:
|
---|
528 | case EXCLAMATIONMARK:
|
---|
529 | // prohibited tokens (used as a multiplier start)
|
---|
530 | break;
|
---|
531 |
|
---|
532 | case LEFTCURLYBRACKET:
|
---|
533 | // LEFTCURLYBRACKET is allowed since mdn/data uses it w/o quoting
|
---|
534 | // check next char isn't a number, because it's likely a disjoined multiplier
|
---|
535 | code = tokenizer.nextCharCode();
|
---|
536 |
|
---|
537 | if (code < 48 || code > 57) {
|
---|
538 | return maybeToken(tokenizer);
|
---|
539 | }
|
---|
540 |
|
---|
541 | break;
|
---|
542 |
|
---|
543 | default:
|
---|
544 | return maybeToken(tokenizer);
|
---|
545 | }
|
---|
546 | }
|
---|
547 |
|
---|
548 | function parse(source) {
|
---|
549 | var tokenizer = new Tokenizer(source);
|
---|
550 | var result = readImplicitGroup(tokenizer);
|
---|
551 |
|
---|
552 | if (tokenizer.pos !== source.length) {
|
---|
553 | tokenizer.error('Unexpected input');
|
---|
554 | }
|
---|
555 |
|
---|
556 | // reduce redundant groups with single group term
|
---|
557 | if (result.terms.length === 1 && result.terms[0].type === 'Group') {
|
---|
558 | result = result.terms[0];
|
---|
559 | }
|
---|
560 |
|
---|
561 | return result;
|
---|
562 | }
|
---|
563 |
|
---|
564 | // warm up parse to elimitate code branches that never execute
|
---|
565 | // fix soft deoptimizations (insufficient type feedback)
|
---|
566 | parse('[a&&<b>#|<\'c\'>*||e() f{2} /,(% g#{1,2} h{2,})]!');
|
---|
567 |
|
---|
568 | module.exports = parse;
|
---|