1 |
9bb1e829
|
cagy
|
/*! https://mths.be/regenerate v1.3.3 by @mathias | MIT license */
|
2 |
|
|
;(function(root) {
|
3 |
|
|
|
4 |
|
|
// Detect free variables `exports`.
|
5 |
|
|
var freeExports = typeof exports == 'object' && exports;
|
6 |
|
|
|
7 |
|
|
// Detect free variable `module`.
|
8 |
|
|
var freeModule = typeof module == 'object' && module &&
|
9 |
|
|
module.exports == freeExports && module;
|
10 |
|
|
|
11 |
|
|
// Detect free variable `global`, from Node.js/io.js or Browserified code,
|
12 |
|
|
// and use it as `root`.
|
13 |
|
|
var freeGlobal = typeof global == 'object' && global;
|
14 |
|
|
if (freeGlobal.global === freeGlobal || freeGlobal.window === freeGlobal) {
|
15 |
|
|
root = freeGlobal;
|
16 |
|
|
}
|
17 |
|
|
|
18 |
|
|
/*--------------------------------------------------------------------------*/
|
19 |
|
|
|
20 |
|
|
var ERRORS = {
|
21 |
|
|
'rangeOrder': 'A range\u2019s `stop` value must be greater than or equal ' +
|
22 |
|
|
'to the `start` value.',
|
23 |
|
|
'codePointRange': 'Invalid code point value. Code points range from ' +
|
24 |
|
|
'U+000000 to U+10FFFF.'
|
25 |
|
|
};
|
26 |
|
|
|
27 |
|
|
// https://mathiasbynens.be/notes/javascript-encoding#surrogate-pairs
|
28 |
|
|
var HIGH_SURROGATE_MIN = 0xD800;
|
29 |
|
|
var HIGH_SURROGATE_MAX = 0xDBFF;
|
30 |
|
|
var LOW_SURROGATE_MIN = 0xDC00;
|
31 |
|
|
var LOW_SURROGATE_MAX = 0xDFFF;
|
32 |
|
|
|
33 |
|
|
// In Regenerate output, `\0` is never preceded by `\` because we sort by
|
34 |
|
|
// code point value, so let’s keep this regular expression simple.
|
35 |
|
|
var regexNull = /\\x00([^0123456789]|$)/g;
|
36 |
|
|
|
37 |
|
|
var object = {};
|
38 |
|
|
var hasOwnProperty = object.hasOwnProperty;
|
39 |
|
|
var extend = function(destination, source) {
|
40 |
|
|
var key;
|
41 |
|
|
for (key in source) {
|
42 |
|
|
if (hasOwnProperty.call(source, key)) {
|
43 |
|
|
destination[key] = source[key];
|
44 |
|
|
}
|
45 |
|
|
}
|
46 |
|
|
return destination;
|
47 |
|
|
};
|
48 |
|
|
|
49 |
|
|
var forEach = function(array, callback) {
|
50 |
|
|
var index = -1;
|
51 |
|
|
var length = array.length;
|
52 |
|
|
while (++index < length) {
|
53 |
|
|
callback(array[index], index);
|
54 |
|
|
}
|
55 |
|
|
};
|
56 |
|
|
|
57 |
|
|
var toString = object.toString;
|
58 |
|
|
var isArray = function(value) {
|
59 |
|
|
return toString.call(value) == '[object Array]';
|
60 |
|
|
};
|
61 |
|
|
var isNumber = function(value) {
|
62 |
|
|
return typeof value == 'number' ||
|
63 |
|
|
toString.call(value) == '[object Number]';
|
64 |
|
|
};
|
65 |
|
|
|
66 |
|
|
// This assumes that `number` is a positive integer that `toString()`s nicely
|
67 |
|
|
// (which is the case for all code point values).
|
68 |
|
|
var zeroes = '0000';
|
69 |
|
|
var pad = function(number, totalCharacters) {
|
70 |
|
|
var string = String(number);
|
71 |
|
|
return string.length < totalCharacters
|
72 |
|
|
? (zeroes + string).slice(-totalCharacters)
|
73 |
|
|
: string;
|
74 |
|
|
};
|
75 |
|
|
|
76 |
|
|
var hex = function(number) {
|
77 |
|
|
return Number(number).toString(16).toUpperCase();
|
78 |
|
|
};
|
79 |
|
|
|
80 |
|
|
var slice = [].slice;
|
81 |
|
|
|
82 |
|
|
/*--------------------------------------------------------------------------*/
|
83 |
|
|
|
84 |
|
|
var dataFromCodePoints = function(codePoints) {
|
85 |
|
|
var index = -1;
|
86 |
|
|
var length = codePoints.length;
|
87 |
|
|
var max = length - 1;
|
88 |
|
|
var result = [];
|
89 |
|
|
var isStart = true;
|
90 |
|
|
var tmp;
|
91 |
|
|
var previous = 0;
|
92 |
|
|
while (++index < length) {
|
93 |
|
|
tmp = codePoints[index];
|
94 |
|
|
if (isStart) {
|
95 |
|
|
result.push(tmp);
|
96 |
|
|
previous = tmp;
|
97 |
|
|
isStart = false;
|
98 |
|
|
} else {
|
99 |
|
|
if (tmp == previous + 1) {
|
100 |
|
|
if (index != max) {
|
101 |
|
|
previous = tmp;
|
102 |
|
|
continue;
|
103 |
|
|
} else {
|
104 |
|
|
isStart = true;
|
105 |
|
|
result.push(tmp + 1);
|
106 |
|
|
}
|
107 |
|
|
} else {
|
108 |
|
|
// End the previous range and start a new one.
|
109 |
|
|
result.push(previous + 1, tmp);
|
110 |
|
|
previous = tmp;
|
111 |
|
|
}
|
112 |
|
|
}
|
113 |
|
|
}
|
114 |
|
|
if (!isStart) {
|
115 |
|
|
result.push(tmp + 1);
|
116 |
|
|
}
|
117 |
|
|
return result;
|
118 |
|
|
};
|
119 |
|
|
|
120 |
|
|
var dataRemove = function(data, codePoint) {
|
121 |
|
|
// Iterate over the data per `(start, end)` pair.
|
122 |
|
|
var index = 0;
|
123 |
|
|
var start;
|
124 |
|
|
var end;
|
125 |
|
|
var length = data.length;
|
126 |
|
|
while (index < length) {
|
127 |
|
|
start = data[index];
|
128 |
|
|
end = data[index + 1];
|
129 |
|
|
if (codePoint >= start && codePoint < end) {
|
130 |
|
|
// Modify this pair.
|
131 |
|
|
if (codePoint == start) {
|
132 |
|
|
if (end == start + 1) {
|
133 |
|
|
// Just remove `start` and `end`.
|
134 |
|
|
data.splice(index, 2);
|
135 |
|
|
return data;
|
136 |
|
|
} else {
|
137 |
|
|
// Just replace `start` with a new value.
|
138 |
|
|
data[index] = codePoint + 1;
|
139 |
|
|
return data;
|
140 |
|
|
}
|
141 |
|
|
} else if (codePoint == end - 1) {
|
142 |
|
|
// Just replace `end` with a new value.
|
143 |
|
|
data[index + 1] = codePoint;
|
144 |
|
|
return data;
|
145 |
|
|
} else {
|
146 |
|
|
// Replace `[start, end]` with `[startA, endA, startB, endB]`.
|
147 |
|
|
data.splice(index, 2, start, codePoint, codePoint + 1, end);
|
148 |
|
|
return data;
|
149 |
|
|
}
|
150 |
|
|
}
|
151 |
|
|
index += 2;
|
152 |
|
|
}
|
153 |
|
|
return data;
|
154 |
|
|
};
|
155 |
|
|
|
156 |
|
|
var dataRemoveRange = function(data, rangeStart, rangeEnd) {
|
157 |
|
|
if (rangeEnd < rangeStart) {
|
158 |
|
|
throw Error(ERRORS.rangeOrder);
|
159 |
|
|
}
|
160 |
|
|
// Iterate over the data per `(start, end)` pair.
|
161 |
|
|
var index = 0;
|
162 |
|
|
var start;
|
163 |
|
|
var end;
|
164 |
|
|
while (index < data.length) {
|
165 |
|
|
start = data[index];
|
166 |
|
|
end = data[index + 1] - 1; // Note: the `- 1` makes `end` inclusive.
|
167 |
|
|
|
168 |
|
|
// Exit as soon as no more matching pairs can be found.
|
169 |
|
|
if (start > rangeEnd) {
|
170 |
|
|
return data;
|
171 |
|
|
}
|
172 |
|
|
|
173 |
|
|
// Check if this range pair is equal to, or forms a subset of, the range
|
174 |
|
|
// to be removed.
|
175 |
|
|
// E.g. we have `[0, 11, 40, 51]` and want to remove 0-10 → `[40, 51]`.
|
176 |
|
|
// E.g. we have `[40, 51]` and want to remove 0-100 → `[]`.
|
177 |
|
|
if (rangeStart <= start && rangeEnd >= end) {
|
178 |
|
|
// Remove this pair.
|
179 |
|
|
data.splice(index, 2);
|
180 |
|
|
continue;
|
181 |
|
|
}
|
182 |
|
|
|
183 |
|
|
// Check if both `rangeStart` and `rangeEnd` are within the bounds of
|
184 |
|
|
// this pair.
|
185 |
|
|
// E.g. we have `[0, 11]` and want to remove 4-6 → `[0, 4, 7, 11]`.
|
186 |
|
|
if (rangeStart >= start && rangeEnd < end) {
|
187 |
|
|
if (rangeStart == start) {
|
188 |
|
|
// Replace `[start, end]` with `[startB, endB]`.
|
189 |
|
|
data[index] = rangeEnd + 1;
|
190 |
|
|
data[index + 1] = end + 1;
|
191 |
|
|
return data;
|
192 |
|
|
}
|
193 |
|
|
// Replace `[start, end]` with `[startA, endA, startB, endB]`.
|
194 |
|
|
data.splice(index, 2, start, rangeStart, rangeEnd + 1, end + 1);
|
195 |
|
|
return data;
|
196 |
|
|
}
|
197 |
|
|
|
198 |
|
|
// Check if only `rangeStart` is within the bounds of this pair.
|
199 |
|
|
// E.g. we have `[0, 11]` and want to remove 4-20 → `[0, 4]`.
|
200 |
|
|
if (rangeStart >= start && rangeStart <= end) {
|
201 |
|
|
// Replace `end` with `rangeStart`.
|
202 |
|
|
data[index + 1] = rangeStart;
|
203 |
|
|
// Note: we cannot `return` just yet, in case any following pairs still
|
204 |
|
|
// contain matching code points.
|
205 |
|
|
// E.g. we have `[0, 11, 14, 31]` and want to remove 4-20
|
206 |
|
|
// → `[0, 4, 21, 31]`.
|
207 |
|
|
}
|
208 |
|
|
|
209 |
|
|
// Check if only `rangeEnd` is within the bounds of this pair.
|
210 |
|
|
// E.g. we have `[14, 31]` and want to remove 4-20 → `[21, 31]`.
|
211 |
|
|
else if (rangeEnd >= start && rangeEnd <= end) {
|
212 |
|
|
// Just replace `start`.
|
213 |
|
|
data[index] = rangeEnd + 1;
|
214 |
|
|
return data;
|
215 |
|
|
}
|
216 |
|
|
|
217 |
|
|
index += 2;
|
218 |
|
|
}
|
219 |
|
|
return data;
|
220 |
|
|
};
|
221 |
|
|
|
222 |
|
|
var dataAdd = function(data, codePoint) {
|
223 |
|
|
// Iterate over the data per `(start, end)` pair.
|
224 |
|
|
var index = 0;
|
225 |
|
|
var start;
|
226 |
|
|
var end;
|
227 |
|
|
var lastIndex = null;
|
228 |
|
|
var length = data.length;
|
229 |
|
|
if (codePoint < 0x0 || codePoint > 0x10FFFF) {
|
230 |
|
|
throw RangeError(ERRORS.codePointRange);
|
231 |
|
|
}
|
232 |
|
|
while (index < length) {
|
233 |
|
|
start = data[index];
|
234 |
|
|
end = data[index + 1];
|
235 |
|
|
|
236 |
|
|
// Check if the code point is already in the set.
|
237 |
|
|
if (codePoint >= start && codePoint < end) {
|
238 |
|
|
return data;
|
239 |
|
|
}
|
240 |
|
|
|
241 |
|
|
if (codePoint == start - 1) {
|
242 |
|
|
// Just replace `start` with a new value.
|
243 |
|
|
data[index] = codePoint;
|
244 |
|
|
return data;
|
245 |
|
|
}
|
246 |
|
|
|
247 |
|
|
// At this point, if `start` is `greater` than `codePoint`, insert a new
|
248 |
|
|
// `[start, end]` pair before the current pair, or after the current pair
|
249 |
|
|
// if there is a known `lastIndex`.
|
250 |
|
|
if (start > codePoint) {
|
251 |
|
|
data.splice(
|
252 |
|
|
lastIndex != null ? lastIndex + 2 : 0,
|
253 |
|
|
0,
|
254 |
|
|
codePoint,
|
255 |
|
|
codePoint + 1
|
256 |
|
|
);
|
257 |
|
|
return data;
|
258 |
|
|
}
|
259 |
|
|
|
260 |
|
|
if (codePoint == end) {
|
261 |
|
|
// Check if adding this code point causes two separate ranges to become
|
262 |
|
|
// a single range, e.g. `dataAdd([0, 4, 5, 10], 4)` → `[0, 10]`.
|
263 |
|
|
if (codePoint + 1 == data[index + 2]) {
|
264 |
|
|
data.splice(index, 4, start, data[index + 3]);
|
265 |
|
|
return data;
|
266 |
|
|
}
|
267 |
|
|
// Else, just replace `end` with a new value.
|
268 |
|
|
data[index + 1] = codePoint + 1;
|
269 |
|
|
return data;
|
270 |
|
|
}
|
271 |
|
|
lastIndex = index;
|
272 |
|
|
index += 2;
|
273 |
|
|
}
|
274 |
|
|
// The loop has finished; add the new pair to the end of the data set.
|
275 |
|
|
data.push(codePoint, codePoint + 1);
|
276 |
|
|
return data;
|
277 |
|
|
};
|
278 |
|
|
|
279 |
|
|
var dataAddData = function(dataA, dataB) {
|
280 |
|
|
// Iterate over the data per `(start, end)` pair.
|
281 |
|
|
var index = 0;
|
282 |
|
|
var start;
|
283 |
|
|
var end;
|
284 |
|
|
var data = dataA.slice();
|
285 |
|
|
var length = dataB.length;
|
286 |
|
|
while (index < length) {
|
287 |
|
|
start = dataB[index];
|
288 |
|
|
end = dataB[index + 1] - 1;
|
289 |
|
|
if (start == end) {
|
290 |
|
|
data = dataAdd(data, start);
|
291 |
|
|
} else {
|
292 |
|
|
data = dataAddRange(data, start, end);
|
293 |
|
|
}
|
294 |
|
|
index += 2;
|
295 |
|
|
}
|
296 |
|
|
return data;
|
297 |
|
|
};
|
298 |
|
|
|
299 |
|
|
var dataRemoveData = function(dataA, dataB) {
|
300 |
|
|
// Iterate over the data per `(start, end)` pair.
|
301 |
|
|
var index = 0;
|
302 |
|
|
var start;
|
303 |
|
|
var end;
|
304 |
|
|
var data = dataA.slice();
|
305 |
|
|
var length = dataB.length;
|
306 |
|
|
while (index < length) {
|
307 |
|
|
start = dataB[index];
|
308 |
|
|
end = dataB[index + 1] - 1;
|
309 |
|
|
if (start == end) {
|
310 |
|
|
data = dataRemove(data, start);
|
311 |
|
|
} else {
|
312 |
|
|
data = dataRemoveRange(data, start, end);
|
313 |
|
|
}
|
314 |
|
|
index += 2;
|
315 |
|
|
}
|
316 |
|
|
return data;
|
317 |
|
|
};
|
318 |
|
|
|
319 |
|
|
var dataAddRange = function(data, rangeStart, rangeEnd) {
|
320 |
|
|
if (rangeEnd < rangeStart) {
|
321 |
|
|
throw Error(ERRORS.rangeOrder);
|
322 |
|
|
}
|
323 |
|
|
if (
|
324 |
|
|
rangeStart < 0x0 || rangeStart > 0x10FFFF ||
|
325 |
|
|
rangeEnd < 0x0 || rangeEnd > 0x10FFFF
|
326 |
|
|
) {
|
327 |
|
|
throw RangeError(ERRORS.codePointRange);
|
328 |
|
|
}
|
329 |
|
|
// Iterate over the data per `(start, end)` pair.
|
330 |
|
|
var index = 0;
|
331 |
|
|
var start;
|
332 |
|
|
var end;
|
333 |
|
|
var added = false;
|
334 |
|
|
var length = data.length;
|
335 |
|
|
while (index < length) {
|
336 |
|
|
start = data[index];
|
337 |
|
|
end = data[index + 1];
|
338 |
|
|
|
339 |
|
|
if (added) {
|
340 |
|
|
// The range has already been added to the set; at this point, we just
|
341 |
|
|
// need to get rid of the following ranges in case they overlap.
|
342 |
|
|
|
343 |
|
|
// Check if this range can be combined with the previous range.
|
344 |
|
|
if (start == rangeEnd + 1) {
|
345 |
|
|
data.splice(index - 1, 2);
|
346 |
|
|
return data;
|
347 |
|
|
}
|
348 |
|
|
|
349 |
|
|
// Exit as soon as no more possibly overlapping pairs can be found.
|
350 |
|
|
if (start > rangeEnd) {
|
351 |
|
|
return data;
|
352 |
|
|
}
|
353 |
|
|
|
354 |
|
|
// E.g. `[0, 11, 12, 16]` and we’ve added 5-15, so we now have
|
355 |
|
|
// `[0, 16, 12, 16]`. Remove the `12,16` part, as it lies within the
|
356 |
|
|
// `0,16` range that was previously added.
|
357 |
|
|
if (start >= rangeStart && start <= rangeEnd) {
|
358 |
|
|
// `start` lies within the range that was previously added.
|
359 |
|
|
|
360 |
|
|
if (end > rangeStart && end - 1 <= rangeEnd) {
|
361 |
|
|
// `end` lies within the range that was previously added as well,
|
362 |
|
|
// so remove this pair.
|
363 |
|
|
data.splice(index, 2);
|
364 |
|
|
index -= 2;
|
365 |
|
|
// Note: we cannot `return` just yet, as there may still be other
|
366 |
|
|
// overlapping pairs.
|
367 |
|
|
} else {
|
368 |
|
|
// `start` lies within the range that was previously added, but
|
369 |
|
|
// `end` doesn’t. E.g. `[0, 11, 12, 31]` and we’ve added 5-15, so
|
370 |
|
|
// now we have `[0, 16, 12, 31]`. This must be written as `[0, 31]`.
|
371 |
|
|
// Remove the previously added `end` and the current `start`.
|
372 |
|
|
data.splice(index - 1, 2);
|
373 |
|
|
index -= 2;
|
374 |
|
|
}
|
375 |
|
|
|
376 |
|
|
// Note: we cannot return yet.
|
377 |
|
|
}
|
378 |
|
|
|
379 |
|
|
}
|
380 |
|
|
|
381 |
|
|
else if (start == rangeEnd + 1) {
|
382 |
|
|
data[index] = rangeStart;
|
383 |
|
|
return data;
|
384 |
|
|
}
|
385 |
|
|
|
386 |
|
|
// Check if a new pair must be inserted *before* the current one.
|
387 |
|
|
else if (start > rangeEnd) {
|
388 |
|
|
data.splice(index, 0, rangeStart, rangeEnd + 1);
|
389 |
|
|
return data;
|
390 |
|
|
}
|
391 |
|
|
|
392 |
|
|
else if (rangeStart >= start && rangeStart < end && rangeEnd + 1 <= end) {
|
393 |
|
|
// The new range lies entirely within an existing range pair. No action
|
394 |
|
|
// needed.
|
395 |
|
|
return data;
|
396 |
|
|
}
|
397 |
|
|
|
398 |
|
|
else if (
|
399 |
|
|
// E.g. `[0, 11]` and you add 5-15 → `[0, 16]`.
|
400 |
|
|
(rangeStart >= start && rangeStart < end) ||
|
401 |
|
|
// E.g. `[0, 3]` and you add 3-6 → `[0, 7]`.
|
402 |
|
|
end == rangeStart
|
403 |
|
|
) {
|
404 |
|
|
// Replace `end` with the new value.
|
405 |
|
|
data[index + 1] = rangeEnd + 1;
|
406 |
|
|
// Make sure the next range pair doesn’t overlap, e.g. `[0, 11, 12, 14]`
|
407 |
|
|
// and you add 5-15 → `[0, 16]`, i.e. remove the `12,14` part.
|
408 |
|
|
added = true;
|
409 |
|
|
// Note: we cannot `return` just yet.
|
410 |
|
|
}
|
411 |
|
|
|
412 |
|
|
else if (rangeStart <= start && rangeEnd + 1 >= end) {
|
413 |
|
|
// The new range is a superset of the old range.
|
414 |
|
|
data[index] = rangeStart;
|
415 |
|
|
data[index + 1] = rangeEnd + 1;
|
416 |
|
|
added = true;
|
417 |
|
|
}
|
418 |
|
|
|
419 |
|
|
index += 2;
|
420 |
|
|
}
|
421 |
|
|
// The loop has finished without doing anything; add the new pair to the end
|
422 |
|
|
// of the data set.
|
423 |
|
|
if (!added) {
|
424 |
|
|
data.push(rangeStart, rangeEnd + 1);
|
425 |
|
|
}
|
426 |
|
|
return data;
|
427 |
|
|
};
|
428 |
|
|
|
429 |
|
|
var dataContains = function(data, codePoint) {
|
430 |
|
|
var index = 0;
|
431 |
|
|
var length = data.length;
|
432 |
|
|
// Exit early if `codePoint` is not within `data`’s overall range.
|
433 |
|
|
var start = data[index];
|
434 |
|
|
var end = data[length - 1];
|
435 |
|
|
if (length >= 2) {
|
436 |
|
|
if (codePoint < start || codePoint > end) {
|
437 |
|
|
return false;
|
438 |
|
|
}
|
439 |
|
|
}
|
440 |
|
|
// Iterate over the data per `(start, end)` pair.
|
441 |
|
|
while (index < length) {
|
442 |
|
|
start = data[index];
|
443 |
|
|
end = data[index + 1];
|
444 |
|
|
if (codePoint >= start && codePoint < end) {
|
445 |
|
|
return true;
|
446 |
|
|
}
|
447 |
|
|
index += 2;
|
448 |
|
|
}
|
449 |
|
|
return false;
|
450 |
|
|
};
|
451 |
|
|
|
452 |
|
|
var dataIntersection = function(data, codePoints) {
|
453 |
|
|
var index = 0;
|
454 |
|
|
var length = codePoints.length;
|
455 |
|
|
var codePoint;
|
456 |
|
|
var result = [];
|
457 |
|
|
while (index < length) {
|
458 |
|
|
codePoint = codePoints[index];
|
459 |
|
|
if (dataContains(data, codePoint)) {
|
460 |
|
|
result.push(codePoint);
|
461 |
|
|
}
|
462 |
|
|
++index;
|
463 |
|
|
}
|
464 |
|
|
return dataFromCodePoints(result);
|
465 |
|
|
};
|
466 |
|
|
|
467 |
|
|
var dataIsEmpty = function(data) {
|
468 |
|
|
return !data.length;
|
469 |
|
|
};
|
470 |
|
|
|
471 |
|
|
var dataIsSingleton = function(data) {
|
472 |
|
|
// Check if the set only represents a single code point.
|
473 |
|
|
return data.length == 2 && data[0] + 1 == data[1];
|
474 |
|
|
};
|
475 |
|
|
|
476 |
|
|
var dataToArray = function(data) {
|
477 |
|
|
// Iterate over the data per `(start, end)` pair.
|
478 |
|
|
var index = 0;
|
479 |
|
|
var start;
|
480 |
|
|
var end;
|
481 |
|
|
var result = [];
|
482 |
|
|
var length = data.length;
|
483 |
|
|
while (index < length) {
|
484 |
|
|
start = data[index];
|
485 |
|
|
end = data[index + 1];
|
486 |
|
|
while (start < end) {
|
487 |
|
|
result.push(start);
|
488 |
|
|
++start;
|
489 |
|
|
}
|
490 |
|
|
index += 2;
|
491 |
|
|
}
|
492 |
|
|
return result;
|
493 |
|
|
};
|
494 |
|
|
|
495 |
|
|
/*--------------------------------------------------------------------------*/
|
496 |
|
|
|
497 |
|
|
// https://mathiasbynens.be/notes/javascript-encoding#surrogate-formulae
|
498 |
|
|
var floor = Math.floor;
|
499 |
|
|
var highSurrogate = function(codePoint) {
|
500 |
|
|
return parseInt(
|
501 |
|
|
floor((codePoint - 0x10000) / 0x400) + HIGH_SURROGATE_MIN,
|
502 |
|
|
10
|
503 |
|
|
);
|
504 |
|
|
};
|
505 |
|
|
|
506 |
|
|
var lowSurrogate = function(codePoint) {
|
507 |
|
|
return parseInt(
|
508 |
|
|
(codePoint - 0x10000) % 0x400 + LOW_SURROGATE_MIN,
|
509 |
|
|
10
|
510 |
|
|
);
|
511 |
|
|
};
|
512 |
|
|
|
513 |
|
|
var stringFromCharCode = String.fromCharCode;
|
514 |
|
|
var codePointToString = function(codePoint) {
|
515 |
|
|
var string;
|
516 |
|
|
// https://mathiasbynens.be/notes/javascript-escapes#single
|
517 |
|
|
// Note: the `\b` escape sequence for U+0008 BACKSPACE in strings has a
|
518 |
|
|
// different meaning in regular expressions (word boundary), so it cannot
|
519 |
|
|
// be used here.
|
520 |
|
|
if (codePoint == 0x09) {
|
521 |
|
|
string = '\\t';
|
522 |
|
|
}
|
523 |
|
|
// Note: IE < 9 treats `'\v'` as `'v'`, so avoid using it.
|
524 |
|
|
// else if (codePoint == 0x0B) {
|
525 |
|
|
// string = '\\v';
|
526 |
|
|
// }
|
527 |
|
|
else if (codePoint == 0x0A) {
|
528 |
|
|
string = '\\n';
|
529 |
|
|
}
|
530 |
|
|
else if (codePoint == 0x0C) {
|
531 |
|
|
string = '\\f';
|
532 |
|
|
}
|
533 |
|
|
else if (codePoint == 0x0D) {
|
534 |
|
|
string = '\\r';
|
535 |
|
|
}
|
536 |
|
|
else if (codePoint == 0x2D) {
|
537 |
|
|
// https://mathiasbynens.be/notes/javascript-escapes#hexadecimal
|
538 |
|
|
// Note: `-` (U+002D HYPHEN-MINUS) is escaped in this way rather
|
539 |
|
|
// than by backslash-escaping, in case the output is used outside
|
540 |
|
|
// of a character class in a `u` RegExp. /\-/u throws, but
|
541 |
|
|
// /\x2D/u is fine.
|
542 |
|
|
string = '\\x2D';
|
543 |
|
|
}
|
544 |
|
|
else if (codePoint == 0x5C) {
|
545 |
|
|
string = '\\\\';
|
546 |
|
|
}
|
547 |
|
|
else if (
|
548 |
|
|
codePoint == 0x24 ||
|
549 |
|
|
(codePoint >= 0x28 && codePoint <= 0x2B) ||
|
550 |
|
|
codePoint == 0x2E || codePoint == 0x2F ||
|
551 |
|
|
codePoint == 0x3F ||
|
552 |
|
|
(codePoint >= 0x5B && codePoint <= 0x5E) ||
|
553 |
|
|
(codePoint >= 0x7B && codePoint <= 0x7D)
|
554 |
|
|
) {
|
555 |
|
|
// The code point maps to an unsafe printable ASCII character;
|
556 |
|
|
// backslash-escape it. Here’s the list of those symbols:
|
557 |
|
|
//
|
558 |
|
|
// $()*+./?[\]^{|}
|
559 |
|
|
//
|
560 |
|
|
// This matches SyntaxCharacters as well as `/` (U+002F SOLIDUS).
|
561 |
|
|
// https://tc39.github.io/ecma262/#prod-SyntaxCharacter
|
562 |
|
|
string = '\\' + stringFromCharCode(codePoint);
|
563 |
|
|
}
|
564 |
|
|
else if (codePoint >= 0x20 && codePoint <= 0x7E) {
|
565 |
|
|
// The code point maps to one of these printable ASCII symbols
|
566 |
|
|
// (including the space character):
|
567 |
|
|
//
|
568 |
|
|
// !"#%&',/0123456789:;<=>@ABCDEFGHIJKLMNO
|
569 |
|
|
// PQRSTUVWXYZ_`abcdefghijklmnopqrstuvwxyz~
|
570 |
|
|
//
|
571 |
|
|
// These can safely be used directly.
|
572 |
|
|
string = stringFromCharCode(codePoint);
|
573 |
|
|
}
|
574 |
|
|
else if (codePoint <= 0xFF) {
|
575 |
|
|
string = '\\x' + pad(hex(codePoint), 2);
|
576 |
|
|
}
|
577 |
|
|
else { // `codePoint <= 0xFFFF` holds true.
|
578 |
|
|
// https://mathiasbynens.be/notes/javascript-escapes#unicode
|
579 |
|
|
string = '\\u' + pad(hex(codePoint), 4);
|
580 |
|
|
}
|
581 |
|
|
|
582 |
|
|
// There’s no need to account for astral symbols / surrogate pairs here,
|
583 |
|
|
// since `codePointToString` is private and only used for BMP code points.
|
584 |
|
|
// But if that’s what you need, just add an `else` block with this code:
|
585 |
|
|
//
|
586 |
|
|
// string = '\\u' + pad(hex(highSurrogate(codePoint)), 4)
|
587 |
|
|
// + '\\u' + pad(hex(lowSurrogate(codePoint)), 4);
|
588 |
|
|
|
589 |
|
|
return string;
|
590 |
|
|
};
|
591 |
|
|
|
592 |
|
|
var codePointToStringUnicode = function(codePoint) {
|
593 |
|
|
if (codePoint <= 0xFFFF) {
|
594 |
|
|
return codePointToString(codePoint);
|
595 |
|
|
}
|
596 |
|
|
return '\\u{' + codePoint.toString(16).toUpperCase() + '}';
|
597 |
|
|
};
|
598 |
|
|
|
599 |
|
|
var symbolToCodePoint = function(symbol) {
|
600 |
|
|
var length = symbol.length;
|
601 |
|
|
var first = symbol.charCodeAt(0);
|
602 |
|
|
var second;
|
603 |
|
|
if (
|
604 |
|
|
first >= HIGH_SURROGATE_MIN && first <= HIGH_SURROGATE_MAX &&
|
605 |
|
|
length > 1 // There is a next code unit.
|
606 |
|
|
) {
|
607 |
|
|
// `first` is a high surrogate, and there is a next character. Assume
|
608 |
|
|
// it’s a low surrogate (else it’s invalid usage of Regenerate anyway).
|
609 |
|
|
second = symbol.charCodeAt(1);
|
610 |
|
|
// https://mathiasbynens.be/notes/javascript-encoding#surrogate-formulae
|
611 |
|
|
return (first - HIGH_SURROGATE_MIN) * 0x400 +
|
612 |
|
|
second - LOW_SURROGATE_MIN + 0x10000;
|
613 |
|
|
}
|
614 |
|
|
return first;
|
615 |
|
|
};
|
616 |
|
|
|
617 |
|
|
var createBMPCharacterClasses = function(data) {
|
618 |
|
|
// Iterate over the data per `(start, end)` pair.
|
619 |
|
|
var result = '';
|
620 |
|
|
var index = 0;
|
621 |
|
|
var start;
|
622 |
|
|
var end;
|
623 |
|
|
var length = data.length;
|
624 |
|
|
if (dataIsSingleton(data)) {
|
625 |
|
|
return codePointToString(data[0]);
|
626 |
|
|
}
|
627 |
|
|
while (index < length) {
|
628 |
|
|
start = data[index];
|
629 |
|
|
end = data[index + 1] - 1; // Note: the `- 1` makes `end` inclusive.
|
630 |
|
|
if (start == end) {
|
631 |
|
|
result += codePointToString(start);
|
632 |
|
|
} else if (start + 1 == end) {
|
633 |
|
|
result += codePointToString(start) + codePointToString(end);
|
634 |
|
|
} else {
|
635 |
|
|
result += codePointToString(start) + '-' + codePointToString(end);
|
636 |
|
|
}
|
637 |
|
|
index += 2;
|
638 |
|
|
}
|
639 |
|
|
return '[' + result + ']';
|
640 |
|
|
};
|
641 |
|
|
|
642 |
|
|
var createUnicodeCharacterClasses = function(data) {
|
643 |
|
|
// Iterate over the data per `(start, end)` pair.
|
644 |
|
|
var result = '';
|
645 |
|
|
var index = 0;
|
646 |
|
|
var start;
|
647 |
|
|
var end;
|
648 |
|
|
var length = data.length;
|
649 |
|
|
if (dataIsSingleton(data)) {
|
650 |
|
|
return codePointToStringUnicode(data[0]);
|
651 |
|
|
}
|
652 |
|
|
while (index < length) {
|
653 |
|
|
start = data[index];
|
654 |
|
|
end = data[index + 1] - 1; // Note: the `- 1` makes `end` inclusive.
|
655 |
|
|
if (start == end) {
|
656 |
|
|
result += codePointToStringUnicode(start);
|
657 |
|
|
} else if (start + 1 == end) {
|
658 |
|
|
result += codePointToStringUnicode(start) + codePointToStringUnicode(end);
|
659 |
|
|
} else {
|
660 |
|
|
result += codePointToStringUnicode(start) + '-' + codePointToStringUnicode(end);
|
661 |
|
|
}
|
662 |
|
|
index += 2;
|
663 |
|
|
}
|
664 |
|
|
return '[' + result + ']';
|
665 |
|
|
};
|
666 |
|
|
|
667 |
|
|
var splitAtBMP = function(data) {
|
668 |
|
|
// Iterate over the data per `(start, end)` pair.
|
669 |
|
|
var loneHighSurrogates = [];
|
670 |
|
|
var loneLowSurrogates = [];
|
671 |
|
|
var bmp = [];
|
672 |
|
|
var astral = [];
|
673 |
|
|
var index = 0;
|
674 |
|
|
var start;
|
675 |
|
|
var end;
|
676 |
|
|
var length = data.length;
|
677 |
|
|
while (index < length) {
|
678 |
|
|
start = data[index];
|
679 |
|
|
end = data[index + 1] - 1; // Note: the `- 1` makes `end` inclusive.
|
680 |
|
|
|
681 |
|
|
if (start < HIGH_SURROGATE_MIN) {
|
682 |
|
|
|
683 |
|
|
// The range starts and ends before the high surrogate range.
|
684 |
|
|
// E.g. (0, 0x10).
|
685 |
|
|
if (end < HIGH_SURROGATE_MIN) {
|
686 |
|
|
bmp.push(start, end + 1);
|
687 |
|
|
}
|
688 |
|
|
|
689 |
|
|
// The range starts before the high surrogate range and ends within it.
|
690 |
|
|
// E.g. (0, 0xD855).
|
691 |
|
|
if (end >= HIGH_SURROGATE_MIN && end <= HIGH_SURROGATE_MAX) {
|
692 |
|
|
bmp.push(start, HIGH_SURROGATE_MIN);
|
693 |
|
|
loneHighSurrogates.push(HIGH_SURROGATE_MIN, end + 1);
|
694 |
|
|
}
|
695 |
|
|
|
696 |
|
|
// The range starts before the high surrogate range and ends in the low
|
697 |
|
|
// surrogate range. E.g. (0, 0xDCFF).
|
698 |
|
|
if (end >= LOW_SURROGATE_MIN && end <= LOW_SURROGATE_MAX) {
|
699 |
|
|
bmp.push(start, HIGH_SURROGATE_MIN);
|
700 |
|
|
loneHighSurrogates.push(HIGH_SURROGATE_MIN, HIGH_SURROGATE_MAX + 1);
|
701 |
|
|
loneLowSurrogates.push(LOW_SURROGATE_MIN, end + 1);
|
702 |
|
|
}
|
703 |
|
|
|
704 |
|
|
// The range starts before the high surrogate range and ends after the
|
705 |
|
|
// low surrogate range. E.g. (0, 0x10FFFF).
|
706 |
|
|
if (end > LOW_SURROGATE_MAX) {
|
707 |
|
|
bmp.push(start, HIGH_SURROGATE_MIN);
|
708 |
|
|
loneHighSurrogates.push(HIGH_SURROGATE_MIN, HIGH_SURROGATE_MAX + 1);
|
709 |
|
|
loneLowSurrogates.push(LOW_SURROGATE_MIN, LOW_SURROGATE_MAX + 1);
|
710 |
|
|
if (end <= 0xFFFF) {
|
711 |
|
|
bmp.push(LOW_SURROGATE_MAX + 1, end + 1);
|
712 |
|
|
} else {
|
713 |
|
|
bmp.push(LOW_SURROGATE_MAX + 1, 0xFFFF + 1);
|
714 |
|
|
astral.push(0xFFFF + 1, end + 1);
|
715 |
|
|
}
|
716 |
|
|
}
|
717 |
|
|
|
718 |
|
|
} else if (start >= HIGH_SURROGATE_MIN && start <= HIGH_SURROGATE_MAX) {
|
719 |
|
|
|
720 |
|
|
// The range starts and ends in the high surrogate range.
|
721 |
|
|
// E.g. (0xD855, 0xD866).
|
722 |
|
|
if (end >= HIGH_SURROGATE_MIN && end <= HIGH_SURROGATE_MAX) {
|
723 |
|
|
loneHighSurrogates.push(start, end + 1);
|
724 |
|
|
}
|
725 |
|
|
|
726 |
|
|
// The range starts in the high surrogate range and ends in the low
|
727 |
|
|
// surrogate range. E.g. (0xD855, 0xDCFF).
|
728 |
|
|
if (end >= LOW_SURROGATE_MIN && end <= LOW_SURROGATE_MAX) {
|
729 |
|
|
loneHighSurrogates.push(start, HIGH_SURROGATE_MAX + 1);
|
730 |
|
|
loneLowSurrogates.push(LOW_SURROGATE_MIN, end + 1);
|
731 |
|
|
}
|
732 |
|
|
|
733 |
|
|
// The range starts in the high surrogate range and ends after the low
|
734 |
|
|
// surrogate range. E.g. (0xD855, 0x10FFFF).
|
735 |
|
|
if (end > LOW_SURROGATE_MAX) {
|
736 |
|
|
loneHighSurrogates.push(start, HIGH_SURROGATE_MAX + 1);
|
737 |
|
|
loneLowSurrogates.push(LOW_SURROGATE_MIN, LOW_SURROGATE_MAX + 1);
|
738 |
|
|
if (end <= 0xFFFF) {
|
739 |
|
|
bmp.push(LOW_SURROGATE_MAX + 1, end + 1);
|
740 |
|
|
} else {
|
741 |
|
|
bmp.push(LOW_SURROGATE_MAX + 1, 0xFFFF + 1);
|
742 |
|
|
astral.push(0xFFFF + 1, end + 1);
|
743 |
|
|
}
|
744 |
|
|
}
|
745 |
|
|
|
746 |
|
|
} else if (start >= LOW_SURROGATE_MIN && start <= LOW_SURROGATE_MAX) {
|
747 |
|
|
|
748 |
|
|
// The range starts and ends in the low surrogate range.
|
749 |
|
|
// E.g. (0xDCFF, 0xDDFF).
|
750 |
|
|
if (end >= LOW_SURROGATE_MIN && end <= LOW_SURROGATE_MAX) {
|
751 |
|
|
loneLowSurrogates.push(start, end + 1);
|
752 |
|
|
}
|
753 |
|
|
|
754 |
|
|
// The range starts in the low surrogate range and ends after the low
|
755 |
|
|
// surrogate range. E.g. (0xDCFF, 0x10FFFF).
|
756 |
|
|
if (end > LOW_SURROGATE_MAX) {
|
757 |
|
|
loneLowSurrogates.push(start, LOW_SURROGATE_MAX + 1);
|
758 |
|
|
if (end <= 0xFFFF) {
|
759 |
|
|
bmp.push(LOW_SURROGATE_MAX + 1, end + 1);
|
760 |
|
|
} else {
|
761 |
|
|
bmp.push(LOW_SURROGATE_MAX + 1, 0xFFFF + 1);
|
762 |
|
|
astral.push(0xFFFF + 1, end + 1);
|
763 |
|
|
}
|
764 |
|
|
}
|
765 |
|
|
|
766 |
|
|
} else if (start > LOW_SURROGATE_MAX && start <= 0xFFFF) {
|
767 |
|
|
|
768 |
|
|
// The range starts and ends after the low surrogate range.
|
769 |
|
|
// E.g. (0xFFAA, 0x10FFFF).
|
770 |
|
|
if (end <= 0xFFFF) {
|
771 |
|
|
bmp.push(start, end + 1);
|
772 |
|
|
} else {
|
773 |
|
|
bmp.push(start, 0xFFFF + 1);
|
774 |
|
|
astral.push(0xFFFF + 1, end + 1);
|
775 |
|
|
}
|
776 |
|
|
|
777 |
|
|
} else {
|
778 |
|
|
|
779 |
|
|
// The range starts and ends in the astral range.
|
780 |
|
|
astral.push(start, end + 1);
|
781 |
|
|
|
782 |
|
|
}
|
783 |
|
|
|
784 |
|
|
index += 2;
|
785 |
|
|
}
|
786 |
|
|
return {
|
787 |
|
|
'loneHighSurrogates': loneHighSurrogates,
|
788 |
|
|
'loneLowSurrogates': loneLowSurrogates,
|
789 |
|
|
'bmp': bmp,
|
790 |
|
|
'astral': astral
|
791 |
|
|
};
|
792 |
|
|
};
|
793 |
|
|
|
794 |
|
|
var optimizeSurrogateMappings = function(surrogateMappings) {
|
795 |
|
|
var result = [];
|
796 |
|
|
var tmpLow = [];
|
797 |
|
|
var addLow = false;
|
798 |
|
|
var mapping;
|
799 |
|
|
var nextMapping;
|
800 |
|
|
var highSurrogates;
|
801 |
|
|
var lowSurrogates;
|
802 |
|
|
var nextHighSurrogates;
|
803 |
|
|
var nextLowSurrogates;
|
804 |
|
|
var index = -1;
|
805 |
|
|
var length = surrogateMappings.length;
|
806 |
|
|
while (++index < length) {
|
807 |
|
|
mapping = surrogateMappings[index];
|
808 |
|
|
nextMapping = surrogateMappings[index + 1];
|
809 |
|
|
if (!nextMapping) {
|
810 |
|
|
result.push(mapping);
|
811 |
|
|
continue;
|
812 |
|
|
}
|
813 |
|
|
highSurrogates = mapping[0];
|
814 |
|
|
lowSurrogates = mapping[1];
|
815 |
|
|
nextHighSurrogates = nextMapping[0];
|
816 |
|
|
nextLowSurrogates = nextMapping[1];
|
817 |
|
|
|
818 |
|
|
// Check for identical high surrogate ranges.
|
819 |
|
|
tmpLow = lowSurrogates;
|
820 |
|
|
while (
|
821 |
|
|
nextHighSurrogates &&
|
822 |
|
|
highSurrogates[0] == nextHighSurrogates[0] &&
|
823 |
|
|
highSurrogates[1] == nextHighSurrogates[1]
|
824 |
|
|
) {
|
825 |
|
|
// Merge with the next item.
|
826 |
|
|
if (dataIsSingleton(nextLowSurrogates)) {
|
827 |
|
|
tmpLow = dataAdd(tmpLow, nextLowSurrogates[0]);
|
828 |
|
|
} else {
|
829 |
|
|
tmpLow = dataAddRange(
|
830 |
|
|
tmpLow,
|
831 |
|
|
nextLowSurrogates[0],
|
832 |
|
|
nextLowSurrogates[1] - 1
|
833 |
|
|
);
|
834 |
|
|
}
|
835 |
|
|
++index;
|
836 |
|
|
mapping = surrogateMappings[index];
|
837 |
|
|
highSurrogates = mapping[0];
|
838 |
|
|
lowSurrogates = mapping[1];
|
839 |
|
|
nextMapping = surrogateMappings[index + 1];
|
840 |
|
|
nextHighSurrogates = nextMapping && nextMapping[0];
|
841 |
|
|
nextLowSurrogates = nextMapping && nextMapping[1];
|
842 |
|
|
addLow = true;
|
843 |
|
|
}
|
844 |
|
|
result.push([
|
845 |
|
|
highSurrogates,
|
846 |
|
|
addLow ? tmpLow : lowSurrogates
|
847 |
|
|
]);
|
848 |
|
|
addLow = false;
|
849 |
|
|
}
|
850 |
|
|
return optimizeByLowSurrogates(result);
|
851 |
|
|
};
|
852 |
|
|
|
853 |
|
|
var optimizeByLowSurrogates = function(surrogateMappings) {
|
854 |
|
|
if (surrogateMappings.length == 1) {
|
855 |
|
|
return surrogateMappings;
|
856 |
|
|
}
|
857 |
|
|
var index = -1;
|
858 |
|
|
var innerIndex = -1;
|
859 |
|
|
while (++index < surrogateMappings.length) {
|
860 |
|
|
var mapping = surrogateMappings[index];
|
861 |
|
|
var lowSurrogates = mapping[1];
|
862 |
|
|
var lowSurrogateStart = lowSurrogates[0];
|
863 |
|
|
var lowSurrogateEnd = lowSurrogates[1];
|
864 |
|
|
innerIndex = index; // Note: the loop starts at the next index.
|
865 |
|
|
while (++innerIndex < surrogateMappings.length) {
|
866 |
|
|
var otherMapping = surrogateMappings[innerIndex];
|
867 |
|
|
var otherLowSurrogates = otherMapping[1];
|
868 |
|
|
var otherLowSurrogateStart = otherLowSurrogates[0];
|
869 |
|
|
var otherLowSurrogateEnd = otherLowSurrogates[1];
|
870 |
|
|
if (
|
871 |
|
|
lowSurrogateStart == otherLowSurrogateStart &&
|
872 |
|
|
lowSurrogateEnd == otherLowSurrogateEnd
|
873 |
|
|
) {
|
874 |
|
|
// Add the code points in the other item to this one.
|
875 |
|
|
if (dataIsSingleton(otherMapping[0])) {
|
876 |
|
|
mapping[0] = dataAdd(mapping[0], otherMapping[0][0]);
|
877 |
|
|
} else {
|
878 |
|
|
mapping[0] = dataAddRange(
|
879 |
|
|
mapping[0],
|
880 |
|
|
otherMapping[0][0],
|
881 |
|
|
otherMapping[0][1] - 1
|
882 |
|
|
);
|
883 |
|
|
}
|
884 |
|
|
// Remove the other, now redundant, item.
|
885 |
|
|
surrogateMappings.splice(innerIndex, 1);
|
886 |
|
|
--innerIndex;
|
887 |
|
|
}
|
888 |
|
|
}
|
889 |
|
|
}
|
890 |
|
|
return surrogateMappings;
|
891 |
|
|
};
|
892 |
|
|
|
893 |
|
|
var surrogateSet = function(data) {
|
894 |
|
|
// Exit early if `data` is an empty set.
|
895 |
|
|
if (!data.length) {
|
896 |
|
|
return [];
|
897 |
|
|
}
|
898 |
|
|
|
899 |
|
|
// Iterate over the data per `(start, end)` pair.
|
900 |
|
|
var index = 0;
|
901 |
|
|
var start;
|
902 |
|
|
var end;
|
903 |
|
|
var startHigh;
|
904 |
|
|
var startLow;
|
905 |
|
|
var endHigh;
|
906 |
|
|
var endLow;
|
907 |
|
|
var surrogateMappings = [];
|
908 |
|
|
var length = data.length;
|
909 |
|
|
while (index < length) {
|
910 |
|
|
start = data[index];
|
911 |
|
|
end = data[index + 1] - 1;
|
912 |
|
|
|
913 |
|
|
startHigh = highSurrogate(start);
|
914 |
|
|
startLow = lowSurrogate(start);
|
915 |
|
|
endHigh = highSurrogate(end);
|
916 |
|
|
endLow = lowSurrogate(end);
|
917 |
|
|
|
918 |
|
|
var startsWithLowestLowSurrogate = startLow == LOW_SURROGATE_MIN;
|
919 |
|
|
var endsWithHighestLowSurrogate = endLow == LOW_SURROGATE_MAX;
|
920 |
|
|
var complete = false;
|
921 |
|
|
|
922 |
|
|
// Append the previous high-surrogate-to-low-surrogate mappings.
|
923 |
|
|
// Step 1: `(startHigh, startLow)` to `(startHigh, LOW_SURROGATE_MAX)`.
|
924 |
|
|
if (
|
925 |
|
|
startHigh == endHigh ||
|
926 |
|
|
startsWithLowestLowSurrogate && endsWithHighestLowSurrogate
|
927 |
|
|
) {
|
928 |
|
|
surrogateMappings.push([
|
929 |
|
|
[startHigh, endHigh + 1],
|
930 |
|
|
[startLow, endLow + 1]
|
931 |
|
|
]);
|
932 |
|
|
complete = true;
|
933 |
|
|
} else {
|
934 |
|
|
surrogateMappings.push([
|
935 |
|
|
[startHigh, startHigh + 1],
|
936 |
|
|
[startLow, LOW_SURROGATE_MAX + 1]
|
937 |
|
|
]);
|
938 |
|
|
}
|
939 |
|
|
|
940 |
|
|
// Step 2: `(startHigh + 1, LOW_SURROGATE_MIN)` to
|
941 |
|
|
// `(endHigh - 1, LOW_SURROGATE_MAX)`.
|
942 |
|
|
if (!complete && startHigh + 1 < endHigh) {
|
943 |
|
|
if (endsWithHighestLowSurrogate) {
|
944 |
|
|
// Combine step 2 and step 3.
|
945 |
|
|
surrogateMappings.push([
|
946 |
|
|
[startHigh + 1, endHigh + 1],
|
947 |
|
|
[LOW_SURROGATE_MIN, endLow + 1]
|
948 |
|
|
]);
|
949 |
|
|
complete = true;
|
950 |
|
|
} else {
|
951 |
|
|
surrogateMappings.push([
|
952 |
|
|
[startHigh + 1, endHigh],
|
953 |
|
|
[LOW_SURROGATE_MIN, LOW_SURROGATE_MAX + 1]
|
954 |
|
|
]);
|
955 |
|
|
}
|
956 |
|
|
}
|
957 |
|
|
|
958 |
|
|
// Step 3. `(endHigh, LOW_SURROGATE_MIN)` to `(endHigh, endLow)`.
|
959 |
|
|
if (!complete) {
|
960 |
|
|
surrogateMappings.push([
|
961 |
|
|
[endHigh, endHigh + 1],
|
962 |
|
|
[LOW_SURROGATE_MIN, endLow + 1]
|
963 |
|
|
]);
|
964 |
|
|
}
|
965 |
|
|
|
966 |
|
|
index += 2;
|
967 |
|
|
}
|
968 |
|
|
|
969 |
|
|
// The format of `surrogateMappings` is as follows:
|
970 |
|
|
//
|
971 |
|
|
// [ surrogateMapping1, surrogateMapping2 ]
|
972 |
|
|
//
|
973 |
|
|
// i.e.:
|
974 |
|
|
//
|
975 |
|
|
// [
|
976 |
|
|
// [ highSurrogates1, lowSurrogates1 ],
|
977 |
|
|
// [ highSurrogates2, lowSurrogates2 ]
|
978 |
|
|
// ]
|
979 |
|
|
return optimizeSurrogateMappings(surrogateMappings);
|
980 |
|
|
};
|
981 |
|
|
|
982 |
|
|
var createSurrogateCharacterClasses = function(surrogateMappings) {
|
983 |
|
|
var result = [];
|
984 |
|
|
forEach(surrogateMappings, function(surrogateMapping) {
|
985 |
|
|
var highSurrogates = surrogateMapping[0];
|
986 |
|
|
var lowSurrogates = surrogateMapping[1];
|
987 |
|
|
result.push(
|
988 |
|
|
createBMPCharacterClasses(highSurrogates) +
|
989 |
|
|
createBMPCharacterClasses(lowSurrogates)
|
990 |
|
|
);
|
991 |
|
|
});
|
992 |
|
|
return result.join('|');
|
993 |
|
|
};
|
994 |
|
|
|
995 |
|
|
var createCharacterClassesFromData = function(data, bmpOnly, hasUnicodeFlag) {
|
996 |
|
|
if (hasUnicodeFlag) {
|
997 |
|
|
return createUnicodeCharacterClasses(data);
|
998 |
|
|
}
|
999 |
|
|
var result = [];
|
1000 |
|
|
|
1001 |
|
|
var parts = splitAtBMP(data);
|
1002 |
|
|
var loneHighSurrogates = parts.loneHighSurrogates;
|
1003 |
|
|
var loneLowSurrogates = parts.loneLowSurrogates;
|
1004 |
|
|
var bmp = parts.bmp;
|
1005 |
|
|
var astral = parts.astral;
|
1006 |
|
|
var hasLoneHighSurrogates = !dataIsEmpty(loneHighSurrogates);
|
1007 |
|
|
var hasLoneLowSurrogates = !dataIsEmpty(loneLowSurrogates);
|
1008 |
|
|
|
1009 |
|
|
var surrogateMappings = surrogateSet(astral);
|
1010 |
|
|
|
1011 |
|
|
if (bmpOnly) {
|
1012 |
|
|
bmp = dataAddData(bmp, loneHighSurrogates);
|
1013 |
|
|
hasLoneHighSurrogates = false;
|
1014 |
|
|
bmp = dataAddData(bmp, loneLowSurrogates);
|
1015 |
|
|
hasLoneLowSurrogates = false;
|
1016 |
|
|
}
|
1017 |
|
|
|
1018 |
|
|
if (!dataIsEmpty(bmp)) {
|
1019 |
|
|
// The data set contains BMP code points that are not high surrogates
|
1020 |
|
|
// needed for astral code points in the set.
|
1021 |
|
|
result.push(createBMPCharacterClasses(bmp));
|
1022 |
|
|
}
|
1023 |
|
|
if (surrogateMappings.length) {
|
1024 |
|
|
// The data set contains astral code points; append character classes
|
1025 |
|
|
// based on their surrogate pairs.
|
1026 |
|
|
result.push(createSurrogateCharacterClasses(surrogateMappings));
|
1027 |
|
|
}
|
1028 |
|
|
// https://gist.github.com/mathiasbynens/bbe7f870208abcfec860
|
1029 |
|
|
if (hasLoneHighSurrogates) {
|
1030 |
|
|
result.push(
|
1031 |
|
|
createBMPCharacterClasses(loneHighSurrogates) +
|
1032 |
|
|
// Make sure the high surrogates aren’t part of a surrogate pair.
|
1033 |
|
|
'(?![\\uDC00-\\uDFFF])'
|
1034 |
|
|
);
|
1035 |
|
|
}
|
1036 |
|
|
if (hasLoneLowSurrogates) {
|
1037 |
|
|
result.push(
|
1038 |
|
|
// It is not possible to accurately assert the low surrogates aren’t
|
1039 |
|
|
// part of a surrogate pair, since JavaScript regular expressions do
|
1040 |
|
|
// not support lookbehind.
|
1041 |
|
|
'(?:[^\\uD800-\\uDBFF]|^)' +
|
1042 |
|
|
createBMPCharacterClasses(loneLowSurrogates)
|
1043 |
|
|
);
|
1044 |
|
|
}
|
1045 |
|
|
return result.join('|');
|
1046 |
|
|
};
|
1047 |
|
|
|
1048 |
|
|
/*--------------------------------------------------------------------------*/
|
1049 |
|
|
|
1050 |
|
|
// `regenerate` can be used as a constructor (and new methods can be added to
|
1051 |
|
|
// its prototype) but also as a regular function, the latter of which is the
|
1052 |
|
|
// documented and most common usage. For that reason, it’s not capitalized.
|
1053 |
|
|
var regenerate = function(value) {
|
1054 |
|
|
if (arguments.length > 1) {
|
1055 |
|
|
value = slice.call(arguments);
|
1056 |
|
|
}
|
1057 |
|
|
if (this instanceof regenerate) {
|
1058 |
|
|
this.data = [];
|
1059 |
|
|
return value ? this.add(value) : this;
|
1060 |
|
|
}
|
1061 |
|
|
return (new regenerate).add(value);
|
1062 |
|
|
};
|
1063 |
|
|
|
1064 |
|
|
regenerate.version = '1.3.3';
|
1065 |
|
|
|
1066 |
|
|
var proto = regenerate.prototype;
|
1067 |
|
|
extend(proto, {
|
1068 |
|
|
'add': function(value) {
|
1069 |
|
|
var $this = this;
|
1070 |
|
|
if (value == null) {
|
1071 |
|
|
return $this;
|
1072 |
|
|
}
|
1073 |
|
|
if (value instanceof regenerate) {
|
1074 |
|
|
// Allow passing other Regenerate instances.
|
1075 |
|
|
$this.data = dataAddData($this.data, value.data);
|
1076 |
|
|
return $this;
|
1077 |
|
|
}
|
1078 |
|
|
if (arguments.length > 1) {
|
1079 |
|
|
value = slice.call(arguments);
|
1080 |
|
|
}
|
1081 |
|
|
if (isArray(value)) {
|
1082 |
|
|
forEach(value, function(item) {
|
1083 |
|
|
$this.add(item);
|
1084 |
|
|
});
|
1085 |
|
|
return $this;
|
1086 |
|
|
}
|
1087 |
|
|
$this.data = dataAdd(
|
1088 |
|
|
$this.data,
|
1089 |
|
|
isNumber(value) ? value : symbolToCodePoint(value)
|
1090 |
|
|
);
|
1091 |
|
|
return $this;
|
1092 |
|
|
},
|
1093 |
|
|
'remove': function(value) {
|
1094 |
|
|
var $this = this;
|
1095 |
|
|
if (value == null) {
|
1096 |
|
|
return $this;
|
1097 |
|
|
}
|
1098 |
|
|
if (value instanceof regenerate) {
|
1099 |
|
|
// Allow passing other Regenerate instances.
|
1100 |
|
|
$this.data = dataRemoveData($this.data, value.data);
|
1101 |
|
|
return $this;
|
1102 |
|
|
}
|
1103 |
|
|
if (arguments.length > 1) {
|
1104 |
|
|
value = slice.call(arguments);
|
1105 |
|
|
}
|
1106 |
|
|
if (isArray(value)) {
|
1107 |
|
|
forEach(value, function(item) {
|
1108 |
|
|
$this.remove(item);
|
1109 |
|
|
});
|
1110 |
|
|
return $this;
|
1111 |
|
|
}
|
1112 |
|
|
$this.data = dataRemove(
|
1113 |
|
|
$this.data,
|
1114 |
|
|
isNumber(value) ? value : symbolToCodePoint(value)
|
1115 |
|
|
);
|
1116 |
|
|
return $this;
|
1117 |
|
|
},
|
1118 |
|
|
'addRange': function(start, end) {
|
1119 |
|
|
var $this = this;
|
1120 |
|
|
$this.data = dataAddRange($this.data,
|
1121 |
|
|
isNumber(start) ? start : symbolToCodePoint(start),
|
1122 |
|
|
isNumber(end) ? end : symbolToCodePoint(end)
|
1123 |
|
|
);
|
1124 |
|
|
return $this;
|
1125 |
|
|
},
|
1126 |
|
|
'removeRange': function(start, end) {
|
1127 |
|
|
var $this = this;
|
1128 |
|
|
var startCodePoint = isNumber(start) ? start : symbolToCodePoint(start);
|
1129 |
|
|
var endCodePoint = isNumber(end) ? end : symbolToCodePoint(end);
|
1130 |
|
|
$this.data = dataRemoveRange(
|
1131 |
|
|
$this.data,
|
1132 |
|
|
startCodePoint,
|
1133 |
|
|
endCodePoint
|
1134 |
|
|
);
|
1135 |
|
|
return $this;
|
1136 |
|
|
},
|
1137 |
|
|
'intersection': function(argument) {
|
1138 |
|
|
var $this = this;
|
1139 |
|
|
// Allow passing other Regenerate instances.
|
1140 |
|
|
// TODO: Optimize this by writing and using `dataIntersectionData()`.
|
1141 |
|
|
var array = argument instanceof regenerate ?
|
1142 |
|
|
dataToArray(argument.data) :
|
1143 |
|
|
argument;
|
1144 |
|
|
$this.data = dataIntersection($this.data, array);
|
1145 |
|
|
return $this;
|
1146 |
|
|
},
|
1147 |
|
|
'contains': function(codePoint) {
|
1148 |
|
|
return dataContains(
|
1149 |
|
|
this.data,
|
1150 |
|
|
isNumber(codePoint) ? codePoint : symbolToCodePoint(codePoint)
|
1151 |
|
|
);
|
1152 |
|
|
},
|
1153 |
|
|
'clone': function() {
|
1154 |
|
|
var set = new regenerate;
|
1155 |
|
|
set.data = this.data.slice(0);
|
1156 |
|
|
return set;
|
1157 |
|
|
},
|
1158 |
|
|
'toString': function(options) {
|
1159 |
|
|
var result = createCharacterClassesFromData(
|
1160 |
|
|
this.data,
|
1161 |
|
|
options ? options.bmpOnly : false,
|
1162 |
|
|
options ? options.hasUnicodeFlag : false
|
1163 |
|
|
);
|
1164 |
|
|
if (!result) {
|
1165 |
|
|
// For an empty set, return something that can be inserted `/here/` to
|
1166 |
|
|
// form a valid regular expression. Avoid `(?:)` since that matches the
|
1167 |
|
|
// empty string.
|
1168 |
|
|
return '[]';
|
1169 |
|
|
}
|
1170 |
|
|
// Use `\0` instead of `\x00` where possible.
|
1171 |
|
|
return result.replace(regexNull, '\\0$1');
|
1172 |
|
|
},
|
1173 |
|
|
'toRegExp': function(flags) {
|
1174 |
|
|
var pattern = this.toString(
|
1175 |
|
|
flags && flags.indexOf('u') != -1 ?
|
1176 |
|
|
{ 'hasUnicodeFlag': true } :
|
1177 |
|
|
null
|
1178 |
|
|
);
|
1179 |
|
|
return RegExp(pattern, flags || '');
|
1180 |
|
|
},
|
1181 |
|
|
'valueOf': function() { // Note: `valueOf` is aliased as `toArray`.
|
1182 |
|
|
return dataToArray(this.data);
|
1183 |
|
|
}
|
1184 |
|
|
});
|
1185 |
|
|
|
1186 |
|
|
proto.toArray = proto.valueOf;
|
1187 |
|
|
|
1188 |
|
|
// Some AMD build optimizers, like r.js, check for specific condition patterns
|
1189 |
|
|
// like the following:
|
1190 |
|
|
if (
|
1191 |
|
|
typeof define == 'function' &&
|
1192 |
|
|
typeof define.amd == 'object' &&
|
1193 |
|
|
define.amd
|
1194 |
|
|
) {
|
1195 |
|
|
define(function() {
|
1196 |
|
|
return regenerate;
|
1197 |
|
|
});
|
1198 |
|
|
} else if (freeExports && !freeExports.nodeType) {
|
1199 |
|
|
if (freeModule) { // in Node.js, io.js, or RingoJS v0.8.0+
|
1200 |
|
|
freeModule.exports = regenerate;
|
1201 |
|
|
} else { // in Narwhal or RingoJS v0.7.0-
|
1202 |
|
|
freeExports.regenerate = regenerate;
|
1203 |
|
|
}
|
1204 |
|
|
} else { // in Rhino or a web browser
|
1205 |
|
|
root.regenerate = regenerate;
|
1206 |
|
|
}
|
1207 |
|
|
|
1208 |
|
|
}(this));
|