Files
OpenMaopaoya/frontend-uni/node_modules/regenerate/regenerate.js
T

1269 lines
36 KiB
JavaScript
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
/*! https://mths.be/regenerate v1.4.2 by @mathias | MIT license */
;(function(root) {
// Detect free variables `exports`.
var freeExports = typeof exports == 'object' && exports;
// Detect free variable `module`.
var freeModule = typeof module == 'object' && module &&
module.exports == freeExports && module;
// Detect free variable `global`, from Node.js/io.js or Browserified code,
// and use it as `root`.
var freeGlobal = typeof global == 'object' && global;
if (freeGlobal.global === freeGlobal || freeGlobal.window === freeGlobal) {
root = freeGlobal;
}
/*--------------------------------------------------------------------------*/
var ERRORS = {
'rangeOrder': 'A range\u2019s `stop` value must be greater than or equal ' +
'to the `start` value.',
'codePointRange': 'Invalid code point value. Code points range from ' +
'U+000000 to U+10FFFF.'
};
// https://mathiasbynens.be/notes/javascript-encoding#surrogate-pairs
var HIGH_SURROGATE_MIN = 0xD800;
var HIGH_SURROGATE_MAX = 0xDBFF;
var LOW_SURROGATE_MIN = 0xDC00;
var LOW_SURROGATE_MAX = 0xDFFF;
// In Regenerate output, `\0` is never preceded by `\` because we sort by
// code point value, so let’s keep this regular expression simple.
var regexNull = /\\x00([^0123456789]|$)/g;
var object = {};
var hasOwnProperty = object.hasOwnProperty;
var extend = function(destination, source) {
var key;
for (key in source) {
if (hasOwnProperty.call(source, key)) {
destination[key] = source[key];
}
}
return destination;
};
var forEach = function(array, callback) {
var index = -1;
var length = array.length;
while (++index < length) {
callback(array[index], index);
}
};
var toString = object.toString;
var isArray = function(value) {
return toString.call(value) == '[object Array]';
};
var isNumber = function(value) {
return typeof value == 'number' ||
toString.call(value) == '[object Number]';
};
// `hexBytes[n]` is the uppercase, zero-padded, two-digit hexadecimal
// representation of `n`, for `0 <= n <= 0xFF`.
var hexBytes = [];
(function() {
var digits = '0123456789ABCDEF';
var index = -1;
while (++index <= 0xFF) {
hexBytes[index] = digits.charAt(index >> 4) + digits.charAt(index & 0xF);
}
}());
// These assume that `number` is an integer in the range the output can
// represent, which is the case for the BMP code points they're used with.
var hex2 = function(number) {
return hexBytes[number];
};
var hex4 = function(number) {
return hexBytes[number >> 8] + hexBytes[number & 0xFF];
};
var slice = [].slice;
/*--------------------------------------------------------------------------*/
var compareNumbers = function(a, b) {
return a - b;
};
// Turns an array of code points into data. The code points may be in any
// order and may contain duplicates. Note: unsorted `codePoints` are sorted
// in place.
var dataFromCodePoints = function(codePoints) {
var length = codePoints.length;
var index = 0;
var start;
var end;
var codePoint;
var result = [];
// Turn runs of consecutive code points into `(start, end)` pairs, skipping
// duplicates. Input is often sorted already, so rather than sorting up
// front, sort and start over only once a code point turns out to be
// smaller than the one before it.
while (index < length) {
// The inner loop only stops at a code point above `end`, so `start` is
// always greater than result[result.length - 1].
start = codePoints[index];
end = start + 1;
while (++index < length && (codePoint = codePoints[index]) <= end) {
if (codePoint == end) {
++end;
} else if (codePoint < end - 1) {
codePoints.sort(compareNumbers);
return dataFromCodePoints(codePoints);
}
}
result.push(start, end);
}
return result;
};
var dataRemove = function(data, codePoint) {
// Iterate over the data per `(start, end)` pair.
var index = 0;
var start;
var end;
var length = data.length;
while (index < length) {
start = data[index];
end = data[index + 1];
if (codePoint >= start && codePoint < end) {
// Modify this pair.
if (codePoint == start) {
if (end == start + 1) {
// Just remove `start` and `end`.
data.splice(index, 2);
return data;
} else {
// Just replace `start` with a new value.
data[index] = codePoint + 1;
return data;
}
} else if (codePoint == end - 1) {
// Just replace `end` with a new value.
data[index + 1] = codePoint;
return data;
} else {
// Replace `[start, end]` with `[startA, endA, startB, endB]`.
data.splice(index, 2, start, codePoint, codePoint + 1, end);
return data;
}
}
index += 2;
}
return data;
};
var dataRemoveRange = function(data, rangeStart, rangeEnd) {
if (rangeEnd < rangeStart) {
throw Error(ERRORS.rangeOrder);
}
var length = data.length;
var rangeEndExclusive = rangeEnd + 1;
var index = 0;
var first;
var start;
var end;
// Skip the pairs that end before the range starts.
while (index < length && data[index + 1] <= rangeStart) {
index += 2;
}
// Find the pairs that overlap the range, i.e. `data[first]` up to (but not
// including) `data[index]`.
first = index;
while (index < length && data[index] < rangeEndExclusive) {
index += 2;
}
if (first == index) {
return data;
}
// Replace the overlapping pairs with a single `splice`, keeping the parts of
// the first and last pair that lie outside the range, if any.
// E.g. we have `[0, 11, 14, 31]` and want to remove 4-20 -> `[0, 4, 21, 31]`.
start = data[first];
end = data[index - 1];
if (start < rangeStart) {
if (end > rangeEndExclusive) {
data.splice(
first, index - first, start, rangeStart, rangeEndExclusive, end
);
} else {
data.splice(first, index - first, start, rangeStart);
}
} else if (end > rangeEndExclusive) {
data.splice(first, index - first, rangeEndExclusive, end);
} else {
data.splice(first, index - first);
}
return data;
};
var dataAdd = function(data, codePoint) {
// Iterate over the data per `(start, end)` pair.
var index = 0;
var start;
var end;
var lastIndex = null;
var length = data.length;
if (codePoint < 0x0 || codePoint > 0x10FFFF) {
throw RangeError(ERRORS.codePointRange);
}
// Fast path: the code point lies at or after the start of the last pair,
// which is the common case when adding code points in ascending order.
if (length && codePoint >= data[length - 2]) {
end = data[length - 1];
if (codePoint > end) {
data.push(codePoint, codePoint + 1);
} else if (codePoint == end) {
data[length - 1] = codePoint + 1;
}
return data;
}
while (index < length) {
start = data[index];
end = data[index + 1];
// Check if the code point is already in the set.
if (codePoint >= start && codePoint < end) {
return data;
}
if (codePoint == start - 1) {
// Just replace `start` with a new value.
data[index] = codePoint;
return data;
}
// At this point, if `start` is `greater` than `codePoint`, insert a new
// `[start, end]` pair before the current pair, or after the current pair
// if there is a known `lastIndex`.
if (start > codePoint) {
data.splice(
lastIndex != null ? lastIndex + 2 : 0,
0,
codePoint,
codePoint + 1
);
return data;
}
if (codePoint == end) {
// Check if adding this code point causes two separate ranges to become
// a single range, e.g. `dataAdd([0, 4, 5, 10], 4)` → `[0, 10]`.
if (codePoint + 1 == data[index + 2]) {
data.splice(index, 4, start, data[index + 3]);
return data;
}
// Else, just replace `end` with a new value.
data[index + 1] = codePoint + 1;
return data;
}
lastIndex = index;
index += 2;
}
// The loop has finished; add the new pair to the end of the data set.
data.push(codePoint, codePoint + 1);
return data;
};
var dataAddRange = function(data, rangeStart, rangeEnd) {
if (rangeEnd < rangeStart) {
throw Error(ERRORS.rangeOrder);
}
if (
rangeStart < 0x0 || rangeStart > 0x10FFFF ||
rangeEnd < 0x0 || rangeEnd > 0x10FFFF
) {
throw RangeError(ERRORS.codePointRange);
}
var length = data.length;
var rangeEndExclusive = rangeEnd + 1;
var index = 0;
var first;
var start;
var end;
// Fast path: the range starts at or after the start of the last pair, which
// is the common case when adding ranges in ascending order.
if (length && rangeStart >= data[length - 2]) {
end = data[length - 1];
if (rangeStart > end) {
data.push(rangeStart, rangeEndExclusive);
} else if (rangeEndExclusive > end) {
// The range overlaps or touches the last pair; extend it.
data[length - 1] = rangeEndExclusive;
}
return data;
}
// Skip the pairs that end before the range starts, without touching it.
while (index < length && data[index + 1] < rangeStart) {
index += 2;
}
// Find the pairs that overlap or touch the range, i.e. `data[first]` up to
// (but not including) `data[index]`.
first = index;
while (index < length && data[index] <= rangeEndExclusive) {
index += 2;
}
if (first == index) {
// Nothing to merge with; insert a new pair.
data.splice(first, 0, rangeStart, rangeEndExclusive);
return data;
}
// Merge the range and the pairs it overlaps or touches into a single pair.
// E.g. `[0, 11, 12, 14, 20, 31]` and you add 5-15 -> `[0, 16, 20, 31]`.
start = data[first] < rangeStart ? data[first] : rangeStart;
end = data[index - 1] > rangeEndExclusive ?
data[index - 1] :
rangeEndExclusive;
if (index - first == 2) {
data[first] = start;
data[first + 1] = end;
} else {
data.splice(first, index - first, start, end);
}
return data;
};
var dataContains = function(data, codePoint) {
// Binary search for the number of `(start, end)` pairs that start at or
// before `codePoint`. `codePoint` is in the set if it lies before the end of
// the last of those pairs.
var length = data.length;
var low = 0;
var high = length >> 1;
var middle;
// Exit early if `codePoint` is not within `data`'s overall range.
if (!length || codePoint < data[0] || codePoint >= data[length - 1]) {
return false;
}
while (low < high) {
middle = (low + high) >> 1;
if (data[middle << 1] <= codePoint) {
low = middle + 1;
} else {
high = middle;
}
}
return low > 0 && codePoint < data[(low << 1) - 1];
};
var dataIntersectionData = function(dataA, dataB) {
var indexA = 0;
var indexB = 0;
var lengthA = dataA.length;
var lengthB = dataB.length;
var startA, endA, startB, endB, start, end;
var result = [];
while (indexA < lengthA && indexB < lengthB) {
startA = dataA[indexA];
endA = dataA[indexA + 1];
startB = dataB[indexB];
endB = dataB[indexB + 1];
start = startA > startB ? startA : startB;
end = endA < endB ? endA : endB;
if (start < end) {
result.push(start, end);
}
if (endA < endB) {
indexA += 2;
} else {
indexB += 2;
}
}
return result;
};
var dataUnion = function(dataA, dataB) {
// Walk both sorted lists of `(start, end)` pairs in lockstep, always taking
// the pair that starts first, and merge it into the previous output pair if
// the two overlap or touch.
var indexA = 0;
var indexB = 0;
var lengthA = dataA.length;
var lengthB = dataB.length;
var start;
var end;
var last;
var result = [];
while (indexA < lengthA || indexB < lengthB) {
if (
indexB >= lengthB ||
(indexA < lengthA && dataA[indexA] <= dataB[indexB])
) {
start = dataA[indexA];
end = dataA[indexA + 1];
indexA += 2;
} else {
start = dataB[indexB];
end = dataB[indexB + 1];
indexB += 2;
}
last = result.length - 1;
if (last > 0 && start <= result[last]) {
if (end > result[last]) {
result[last] = end;
}
} else {
result.push(start, end);
}
}
return result;
};
var dataDifference = function(dataA, dataB) {
// For each pair in `dataA`, cut out the overlapping pairs of `dataB`.
var indexA = 0;
var indexB = 0;
var lengthA = dataA.length;
var lengthB = dataB.length;
var index;
var start;
var end;
var result = [];
while (indexA < lengthA) {
start = dataA[indexA];
end = dataA[indexA + 1];
indexA += 2;
// Skip the pairs of `dataB` that end before this pair starts.
while (indexB < lengthB && dataB[indexB + 1] <= start) {
indexB += 2;
}
index = indexB;
while (start < end && index < lengthB && dataB[index] < end) {
if (dataB[index] > start) {
result.push(start, dataB[index]);
}
if (dataB[index + 1] > start) {
start = dataB[index + 1];
}
index += 2;
}
if (start < end) {
result.push(start, end);
}
}
return result;
};
// Collects the code points in `values` (anything `add` or `remove` accepts in
// an array: code points, symbols, nested arrays, and Regenerate instances)
// and returns them as data. With `validate`, out-of-range code points throw
// like `dataAdd` does; otherwise they’re skipped, as `dataRemove` does.
var dataFromValues = function(values, validate) {
var codePoints = [];
var sets = [];
var collect = function(values) {
var index = -1;
var length = values.length;
var value;
var codePoint;
while (++index < length) {
value = values[index];
// Check for primitive numbers first, as they’re the common case.
if (typeof value == 'number') {
codePoint = value;
} else if (value == null) {
continue;
} else if (value instanceof regenerate) {
sets.push(value.data);
continue;
} else if (isArray(value)) {
collect(value);
continue;
} else {
codePoint = isNumber(value) ? +value : symbolToCodePoint(value);
}
if (codePoint >= 0x0 && codePoint <= 0x10FFFF) {
codePoints.push(codePoint);
} else if (validate) {
throw RangeError(ERRORS.codePointRange);
}
}
};
collect(values);
var data = dataFromCodePoints(codePoints);
var index = -1;
var length = sets.length;
while (++index < length) {
data = dataUnion(data, sets[index]);
}
return data;
};
var dataIsEmpty = function(data) {
return !data.length;
};
var dataIsSingleton = function(data) {
// Check if the set only represents a single code point.
return data.length == 2 && data[0] + 1 == data[1];
};
var dataToArray = function(data) {
// Iterate over the data per `(start, end)` pair.
var index = 0;
var start;
var end;
var result = [];
var length = data.length;
while (index < length) {
start = data[index];
end = data[index + 1];
while (start < end) {
result.push(start);
++start;
}
index += 2;
}
return result;
};
/*--------------------------------------------------------------------------*/
// https://mathiasbynens.be/notes/javascript-encoding#surrogate-formulae
var floor = Math.floor;
var highSurrogate = function(codePoint) {
return parseInt(
floor((codePoint - 0x10000) / 0x400) + HIGH_SURROGATE_MIN,
10
);
};
var lowSurrogate = function(codePoint) {
return parseInt(
(codePoint - 0x10000) % 0x400 + LOW_SURROGATE_MIN,
10
);
};
var stringFromCharCode = String.fromCharCode;
var codePointToString = function(codePoint) {
var string;
// https://mathiasbynens.be/notes/javascript-escapes#single
// Note: the `\b` escape sequence for U+0008 BACKSPACE in strings has a
// different meaning in regular expressions (word boundary), so it cannot
// be used here.
if (codePoint == 0x09) {
string = '\\t';
}
// Note: IE < 9 treats `'\v'` as `'v'`, so avoid using it.
// else if (codePoint == 0x0B) {
// string = '\\v';
// }
else if (codePoint == 0x0A) {
string = '\\n';
}
else if (codePoint == 0x0C) {
string = '\\f';
}
else if (codePoint == 0x0D) {
string = '\\r';
}
else if (codePoint == 0x2D) {
// https://mathiasbynens.be/notes/javascript-escapes#hexadecimal
// Note: `-` (U+002D HYPHEN-MINUS) is escaped in this way rather
// than by backslash-escaping, in case the output is used outside
// of a character class in a `u` RegExp. /\-/u throws, but
// /\x2D/u is fine.
string = '\\x2D';
}
else if (codePoint == 0x5C) {
string = '\\\\';
}
else if (
codePoint == 0x24 ||
(codePoint >= 0x28 && codePoint <= 0x2B) ||
codePoint == 0x2E || codePoint == 0x2F ||
codePoint == 0x3F ||
(codePoint >= 0x5B && codePoint <= 0x5E) ||
(codePoint >= 0x7B && codePoint <= 0x7D)
) {
// The code point maps to an unsafe printable ASCII character;
// backslash-escape it. Here’s the list of those symbols:
//
// $()*+./?[\]^{|}
//
// This matches SyntaxCharacters as well as `/` (U+002F SOLIDUS).
// https://tc39.github.io/ecma262/#prod-SyntaxCharacter
string = '\\' + stringFromCharCode(codePoint);
}
else if (codePoint >= 0x20 && codePoint <= 0x7E) {
// The code point maps to one of these printable ASCII symbols
// (including the space character):
//
// !"#%&',0123456789:;<=>@ABCDEFGHIJKLMNO
// PQRSTUVWXYZ_`abcdefghijklmnopqrstuvwxyz~
//
// These can safely be used directly.
string = stringFromCharCode(codePoint);
}
else if (codePoint <= 0xFF) {
string = '\\x' + hex2(codePoint);
}
else { // `codePoint <= 0xFFFF` holds true.
// https://mathiasbynens.be/notes/javascript-escapes#unicode
string = '\\u' + hex4(codePoint);
}
// There’s no need to account for astral symbols / surrogate pairs here,
// since `codePointToString` is private and only used for BMP code points.
// But if that’s what you need, just add an `else` block with this code:
//
// string = '\\u' + hex4(highSurrogate(codePoint))
// + '\\u' + hex4(lowSurrogate(codePoint));
return string;
};
var codePointToStringUnicode = function(codePoint) {
if (codePoint <= 0xFFFF) {
if (codePoint < HIGH_SURROGATE_MIN || codePoint > LOW_SURROGATE_MAX) {
return codePointToString(codePoint);
}
// With the `u` flag, `\uD834\uDF06` is read as the single code point
// U+1D306, so a high surrogate followed by a low surrogate would no
// longer match either of them. `\u{…}` escapes are never combined.
return '\\u{' + hex4(codePoint) + '}';
}
// Astral code points have five or six hexadecimal digits: their highest
// byte is 0x01 to 0x10, written without its leading zero, if any.
var high = codePoint >> 16;
return '\\u{' +
(high < 0x10 ? hexBytes[high].charAt(1) : '10') +
hexBytes[(codePoint >> 8) & 0xFF] +
hexBytes[codePoint & 0xFF] +
'}';
};
var symbolToCodePoint = function(symbol) {
var length = symbol.length;
var first = symbol.charCodeAt(0);
var second;
if (
first >= HIGH_SURROGATE_MIN && first <= HIGH_SURROGATE_MAX &&
length > 1 // There is a next code unit.
) {
// `first` is a high surrogate, and there is a next character. Assume
// it’s a low surrogate (else it’s invalid usage of Regenerate anyway).
second = symbol.charCodeAt(1);
// https://mathiasbynens.be/notes/javascript-encoding#surrogate-formulae
return (first - HIGH_SURROGATE_MIN) * 0x400 +
second - LOW_SURROGATE_MIN + 0x10000;
}
return first;
};
// Turns a code point or a symbol into a primitive code point, so that
// `Number` objects don’t end up in the data. Primitive numbers are returned
// as is, which keeps them fast to work with.
var toCodePoint = function(value) {
return typeof value == 'number' ? value :
isNumber(value) ? +value : symbolToCodePoint(value);
};
var createBMPCharacterClasses = function(data) {
// Iterate over the data per `(start, end)` pair.
var result = '';
var index = 0;
var start;
var end;
var length = data.length;
if (dataIsSingleton(data)) {
return codePointToString(data[0]);
}
while (index < length) {
start = data[index];
end = data[index + 1] - 1; // Note: the `- 1` makes `end` inclusive.
if (start == end) {
result += codePointToString(start);
} else if (start + 1 == end) {
result += codePointToString(start) + codePointToString(end);
} else {
result += codePointToString(start) + '-' + codePointToString(end);
}
index += 2;
}
return '[' + result + ']';
};
var createUnicodeCharacterClasses = function(data) {
// Iterate over the data per `(start, end)` pair.
var result = '';
var index = 0;
var start;
var end;
var length = data.length;
if (dataIsSingleton(data)) {
return codePointToStringUnicode(data[0]);
}
while (index < length) {
start = data[index];
end = data[index + 1] - 1; // Note: the `- 1` makes `end` inclusive.
if (start == end) {
result += codePointToStringUnicode(start);
} else if (start + 1 == end) {
result += codePointToStringUnicode(start) + codePointToStringUnicode(end);
} else {
result += codePointToStringUnicode(start) + '-' + codePointToStringUnicode(end);
}
index += 2;
}
return '[' + result + ']';
};
var splitAtBMP = function(data) {
// Iterate over the data per `(start, end)` pair.
var loneHighSurrogates = [];
var loneLowSurrogates = [];
var bmp = [];
var astral = [];
var index = 0;
var start;
var end;
var length = data.length;
while (index < length) {
start = data[index];
end = data[index + 1] - 1; // Note: the `- 1` makes `end` inclusive.
if (start < HIGH_SURROGATE_MIN) {
// The range starts and ends before the high surrogate range.
// E.g. (0, 0x10).
if (end < HIGH_SURROGATE_MIN) {
bmp.push(start, end + 1);
}
// The range starts before the high surrogate range and ends within it.
// E.g. (0, 0xD855).
if (end >= HIGH_SURROGATE_MIN && end <= HIGH_SURROGATE_MAX) {
bmp.push(start, HIGH_SURROGATE_MIN);
loneHighSurrogates.push(HIGH_SURROGATE_MIN, end + 1);
}
// The range starts before the high surrogate range and ends in the low
// surrogate range. E.g. (0, 0xDCFF).
if (end >= LOW_SURROGATE_MIN && end <= LOW_SURROGATE_MAX) {
bmp.push(start, HIGH_SURROGATE_MIN);
loneHighSurrogates.push(HIGH_SURROGATE_MIN, HIGH_SURROGATE_MAX + 1);
loneLowSurrogates.push(LOW_SURROGATE_MIN, end + 1);
}
// The range starts before the high surrogate range and ends after the
// low surrogate range. E.g. (0, 0x10FFFF).
if (end > LOW_SURROGATE_MAX) {
bmp.push(start, HIGH_SURROGATE_MIN);
loneHighSurrogates.push(HIGH_SURROGATE_MIN, HIGH_SURROGATE_MAX + 1);
loneLowSurrogates.push(LOW_SURROGATE_MIN, LOW_SURROGATE_MAX + 1);
if (end <= 0xFFFF) {
bmp.push(LOW_SURROGATE_MAX + 1, end + 1);
} else {
bmp.push(LOW_SURROGATE_MAX + 1, 0xFFFF + 1);
astral.push(0xFFFF + 1, end + 1);
}
}
} else if (start >= HIGH_SURROGATE_MIN && start <= HIGH_SURROGATE_MAX) {
// The range starts and ends in the high surrogate range.
// E.g. (0xD855, 0xD866).
if (end >= HIGH_SURROGATE_MIN && end <= HIGH_SURROGATE_MAX) {
loneHighSurrogates.push(start, end + 1);
}
// The range starts in the high surrogate range and ends in the low
// surrogate range. E.g. (0xD855, 0xDCFF).
if (end >= LOW_SURROGATE_MIN && end <= LOW_SURROGATE_MAX) {
loneHighSurrogates.push(start, HIGH_SURROGATE_MAX + 1);
loneLowSurrogates.push(LOW_SURROGATE_MIN, end + 1);
}
// The range starts in the high surrogate range and ends after the low
// surrogate range. E.g. (0xD855, 0x10FFFF).
if (end > LOW_SURROGATE_MAX) {
loneHighSurrogates.push(start, HIGH_SURROGATE_MAX + 1);
loneLowSurrogates.push(LOW_SURROGATE_MIN, LOW_SURROGATE_MAX + 1);
if (end <= 0xFFFF) {
bmp.push(LOW_SURROGATE_MAX + 1, end + 1);
} else {
bmp.push(LOW_SURROGATE_MAX + 1, 0xFFFF + 1);
astral.push(0xFFFF + 1, end + 1);
}
}
} else if (start >= LOW_SURROGATE_MIN && start <= LOW_SURROGATE_MAX) {
// The range starts and ends in the low surrogate range.
// E.g. (0xDCFF, 0xDDFF).
if (end >= LOW_SURROGATE_MIN && end <= LOW_SURROGATE_MAX) {
loneLowSurrogates.push(start, end + 1);
}
// The range starts in the low surrogate range and ends after the low
// surrogate range. E.g. (0xDCFF, 0x10FFFF).
if (end > LOW_SURROGATE_MAX) {
loneLowSurrogates.push(start, LOW_SURROGATE_MAX + 1);
if (end <= 0xFFFF) {
bmp.push(LOW_SURROGATE_MAX + 1, end + 1);
} else {
bmp.push(LOW_SURROGATE_MAX + 1, 0xFFFF + 1);
astral.push(0xFFFF + 1, end + 1);
}
}
} else if (start > LOW_SURROGATE_MAX && start <= 0xFFFF) {
// The range starts and ends after the low surrogate range.
// E.g. (0xFFAA, 0x10FFFF).
if (end <= 0xFFFF) {
bmp.push(start, end + 1);
} else {
bmp.push(start, 0xFFFF + 1);
astral.push(0xFFFF + 1, end + 1);
}
} else {
// The range starts and ends in the astral range.
astral.push(start, end + 1);
}
index += 2;
}
return {
'loneHighSurrogates': loneHighSurrogates,
'loneLowSurrogates': loneLowSurrogates,
'bmp': bmp,
'astral': astral
};
};
var optimizeSurrogateMappings = function(surrogateMappings) {
var result = [];
var tmpLow = [];
var addLow = false;
var mapping;
var nextMapping;
var highSurrogates;
var lowSurrogates;
var nextHighSurrogates;
var nextLowSurrogates;
var index = -1;
var length = surrogateMappings.length;
while (++index < length) {
mapping = surrogateMappings[index];
nextMapping = surrogateMappings[index + 1];
if (!nextMapping) {
result.push(mapping);
continue;
}
highSurrogates = mapping[0];
lowSurrogates = mapping[1];
nextHighSurrogates = nextMapping[0];
nextLowSurrogates = nextMapping[1];
// Check for identical high surrogate ranges.
tmpLow = lowSurrogates;
while (
nextHighSurrogates &&
highSurrogates[0] == nextHighSurrogates[0] &&
highSurrogates[1] == nextHighSurrogates[1]
) {
// Merge with the next item.
if (dataIsSingleton(nextLowSurrogates)) {
tmpLow = dataAdd(tmpLow, nextLowSurrogates[0]);
} else {
tmpLow = dataAddRange(
tmpLow,
nextLowSurrogates[0],
nextLowSurrogates[1] - 1
);
}
++index;
mapping = surrogateMappings[index];
highSurrogates = mapping[0];
lowSurrogates = mapping[1];
nextMapping = surrogateMappings[index + 1];
nextHighSurrogates = nextMapping && nextMapping[0];
nextLowSurrogates = nextMapping && nextMapping[1];
addLow = true;
}
result.push([
highSurrogates,
addLow ? tmpLow : lowSurrogates
]);
addLow = false;
}
return optimizeByLowSurrogates(result);
};
var optimizeByLowSurrogates = function(surrogateMappings) {
if (surrogateMappings.length == 1) {
return surrogateMappings;
}
var index = -1;
var innerIndex = -1;
while (++index < surrogateMappings.length) {
var mapping = surrogateMappings[index];
var lowSurrogates = mapping[1];
if (lowSurrogates.length !== 2) {
// Merge can only be performed when both lowSurrogates.length and
// otherLowSurrogates.length are equal, that is, both are single
// ranges of length 2.
continue;
}
var lowSurrogateStart = lowSurrogates[0];
var lowSurrogateEnd = lowSurrogates[1];
innerIndex = index; // Note: the loop starts at the next index.
while (++innerIndex < surrogateMappings.length) {
var otherMapping = surrogateMappings[innerIndex];
var otherLowSurrogates = otherMapping[1];
var otherLowSurrogateStart = otherLowSurrogates[0];
var otherLowSurrogateEnd = otherLowSurrogates[1];
if (
lowSurrogateStart == otherLowSurrogateStart &&
lowSurrogateEnd == otherLowSurrogateEnd &&
otherLowSurrogates.length === 2
) {
// Add the code points in the other item to this one.
if (dataIsSingleton(otherMapping[0])) {
mapping[0] = dataAdd(mapping[0], otherMapping[0][0]);
} else {
mapping[0] = dataAddRange(
mapping[0],
otherMapping[0][0],
otherMapping[0][1] - 1
);
}
// Remove the other, now redundant, item.
surrogateMappings.splice(innerIndex, 1);
--innerIndex;
}
}
}
return surrogateMappings;
};
var surrogateSet = function(data) {
// Exit early if `data` is an empty set.
if (!data.length) {
return [];
}
// Iterate over the data per `(start, end)` pair.
var index = 0;
var start;
var end;
var startHigh;
var startLow;
var endHigh;
var endLow;
var surrogateMappings = [];
var length = data.length;
while (index < length) {
start = data[index];
end = data[index + 1] - 1;
startHigh = highSurrogate(start);
startLow = lowSurrogate(start);
endHigh = highSurrogate(end);
endLow = lowSurrogate(end);
var startsWithLowestLowSurrogate = startLow == LOW_SURROGATE_MIN;
var endsWithHighestLowSurrogate = endLow == LOW_SURROGATE_MAX;
var complete = false;
// Append the previous high-surrogate-to-low-surrogate mappings.
// Step 1: `(startHigh, startLow)` to `(startHigh, LOW_SURROGATE_MAX)`.
if (
startHigh == endHigh ||
startsWithLowestLowSurrogate && endsWithHighestLowSurrogate
) {
surrogateMappings.push([
[startHigh, endHigh + 1],
[startLow, endLow + 1]
]);
complete = true;
} else {
surrogateMappings.push([
[startHigh, startHigh + 1],
[startLow, LOW_SURROGATE_MAX + 1]
]);
}
// Step 2: `(startHigh + 1, LOW_SURROGATE_MIN)` to
// `(endHigh - 1, LOW_SURROGATE_MAX)`.
if (!complete && startHigh + 1 < endHigh) {
if (endsWithHighestLowSurrogate) {
// Combine step 2 and step 3.
surrogateMappings.push([
[startHigh + 1, endHigh + 1],
[LOW_SURROGATE_MIN, endLow + 1]
]);
complete = true;
} else {
surrogateMappings.push([
[startHigh + 1, endHigh],
[LOW_SURROGATE_MIN, LOW_SURROGATE_MAX + 1]
]);
}
}
// Step 3. `(endHigh, LOW_SURROGATE_MIN)` to `(endHigh, endLow)`.
if (!complete) {
surrogateMappings.push([
[endHigh, endHigh + 1],
[LOW_SURROGATE_MIN, endLow + 1]
]);
}
index += 2;
}
// The format of `surrogateMappings` is as follows:
//
// [ surrogateMapping1, surrogateMapping2 ]
//
// i.e.:
//
// [
// [ highSurrogates1, lowSurrogates1 ],
// [ highSurrogates2, lowSurrogates2 ]
// ]
return optimizeSurrogateMappings(surrogateMappings);
};
var createSurrogateCharacterClasses = function(surrogateMappings) {
var result = [];
forEach(surrogateMappings, function(surrogateMapping) {
var highSurrogates = surrogateMapping[0];
var lowSurrogates = surrogateMapping[1];
result.push(
createBMPCharacterClasses(highSurrogates) +
createBMPCharacterClasses(lowSurrogates)
);
});
return result.join('|');
};
var createCharacterClassesFromData = function(data, bmpOnly, hasUnicodeFlag) {
if (hasUnicodeFlag) {
return createUnicodeCharacterClasses(data);
}
var result = [];
var parts = splitAtBMP(data);
var loneHighSurrogates = parts.loneHighSurrogates;
var loneLowSurrogates = parts.loneLowSurrogates;
var bmp = parts.bmp;
var astral = parts.astral;
var hasLoneHighSurrogates = !dataIsEmpty(loneHighSurrogates);
var hasLoneLowSurrogates = !dataIsEmpty(loneLowSurrogates);
var surrogateMappings = surrogateSet(astral);
if (bmpOnly) {
bmp = dataUnion(bmp, loneHighSurrogates);
hasLoneHighSurrogates = false;
bmp = dataUnion(bmp, loneLowSurrogates);
hasLoneLowSurrogates = false;
}
if (!dataIsEmpty(bmp)) {
// The data set contains BMP code points that are not high surrogates
// needed for astral code points in the set.
result.push(createBMPCharacterClasses(bmp));
}
if (surrogateMappings.length) {
// The data set contains astral code points; append character classes
// based on their surrogate pairs.
result.push(createSurrogateCharacterClasses(surrogateMappings));
}
// https://gist.github.com/mathiasbynens/bbe7f870208abcfec860
if (hasLoneHighSurrogates) {
result.push(
createBMPCharacterClasses(loneHighSurrogates) +
// Make sure the high surrogates aren’t part of a surrogate pair.
'(?![\\uDC00-\\uDFFF])'
);
}
if (hasLoneLowSurrogates) {
result.push(
// It is not possible to accurately assert the low surrogates aren’t
// part of a surrogate pair, since JavaScript regular expressions do
// not support lookbehind.
'(?:[^\\uD800-\\uDBFF]|^)' +
createBMPCharacterClasses(loneLowSurrogates)
);
}
return result.join('|');
};
/*--------------------------------------------------------------------------*/
// `regenerate` can be used as a constructor (and new methods can be added to
// its prototype) but also as a regular function, the latter of which is the
// documented and most common usage. For that reason, it’s not capitalized.
var regenerate = function(value) {
if (arguments.length > 1) {
value = slice.call(arguments);
}
if (this instanceof regenerate) {
this.data = [];
return this.add(value);
}
return (new regenerate).add(value);
};
regenerate.version = '1.4.2';
var proto = regenerate.prototype;
extend(proto, {
'add': function(value) {
var $this = this;
if (arguments.length > 1) {
value = slice.call(arguments);
}
if (value == null) {
return $this;
}
if (value instanceof regenerate) {
// Allow passing other Regenerate instances.
$this.data = dataUnion($this.data, value.data);
return $this;
}
if (isArray(value)) {
$this.data = dataUnion($this.data, dataFromValues(value, true));
return $this;
}
$this.data = dataAdd(
$this.data,
toCodePoint(value)
);
return $this;
},
'remove': function(value) {
var $this = this;
if (arguments.length > 1) {
value = slice.call(arguments);
}
if (value == null) {
return $this;
}
if (value instanceof regenerate) {
// Allow passing other Regenerate instances.
$this.data = dataDifference($this.data, value.data);
return $this;
}
if (isArray(value)) {
$this.data = dataDifference($this.data, dataFromValues(value, false));
return $this;
}
$this.data = dataRemove(
$this.data,
toCodePoint(value)
);
return $this;
},
'addRange': function(start, end) {
var $this = this;
$this.data = dataAddRange($this.data,
toCodePoint(start),
toCodePoint(end)
);
return $this;
},
'removeRange': function(start, end) {
var $this = this;
var startCodePoint = toCodePoint(start);
var endCodePoint = toCodePoint(end);
$this.data = dataRemoveRange(
$this.data,
startCodePoint,
endCodePoint
);
return $this;
},
'intersection': function(argument) {
var $this = this;
if (argument instanceof regenerate) {
// Allow passing other Regenerate instances.
$this.data = dataIntersectionData($this.data, argument.data);
return $this;
}
$this.data = dataIntersectionData(
$this.data,
dataFromValues(argument, false)
);
return $this;
},
'contains': function(codePoint) {
return dataContains(
this.data,
isNumber(codePoint) ? codePoint : symbolToCodePoint(codePoint)
);
},
'clone': function() {
var set = new regenerate;
set.data = this.data.slice(0);
return set;
},
'toString': function(options) {
var result = createCharacterClassesFromData(
this.data,
options ? options.bmpOnly : false,
options ? options.hasUnicodeFlag : false
);
if (!result) {
// For an empty set, return something that can be inserted `/here/` to
// form a valid regular expression. Avoid `(?:)` since that matches the
// empty string.
return '[]';
}
// Use `\0` instead of `\x00` where possible.
return result.replace(regexNull, '\\0$1');
},
'toRegExp': function(flags) {
var pattern = this.toString(
/[uv]/.test(flags || '') ?
{ 'hasUnicodeFlag': true } :
null
);
return RegExp(pattern, flags || '');
},
'valueOf': function() { // Note: `valueOf` is aliased as `toArray`.
return dataToArray(this.data);
}
});
proto.toArray = proto.valueOf;
// Some AMD build optimizers, like r.js, check for specific condition patterns
// like the following:
if (
typeof define == 'function' &&
typeof define.amd == 'object' &&
define.amd
) {
define(function() {
return regenerate;
});
} else if (freeExports && !freeExports.nodeType) {
if (freeModule) { // in Node.js, io.js, or RingoJS v0.8.0+
freeModule.exports = regenerate;
} else { // in Narwhal or RingoJS v0.7.0-
freeExports.regenerate = regenerate;
}
} else { // in Rhino or a web browser
root.regenerate = regenerate;
}
}(this));