@automattic/vip 4.1.4 → 4.1.6

package/node_modules/unicode-segmenter/_grapheme_data.d.ts

@@ -0,0 +1,112 @@
+/**
+ * @typedef {import('./core.js').UnicodeDataEncoding} UnicodeDataEncoding
+ */
+/**
+ * @typedef {0} GC_Any
+ * @typedef {1} GC_CR
+ * @typedef {2} GC_Control
+ * @typedef {3} GC_Extend
+ * @typedef {4} GC_Extended_Pictographic
+ * @typedef {5} GC_L
+ * @typedef {6} GC_LF
+ * @typedef {7} GC_LV
+ * @typedef {8} GC_LVT
+ * @typedef {9} GC_Prepend
+ * @typedef {10} GC_Regional_Indicator
+ * @typedef {11} GC_SpacingMark
+ * @typedef {12} GC_T
+ * @typedef {13} GC_V
+ * @typedef {14} GC_ZWJ
+ * @typedef {(
+ *   | GC_Any
+ *   | GC_CR
+ *   | GC_Control
+ *   | GC_Extend
+ *   | GC_Extended_Pictographic
+ *   | GC_L
+ *   | GC_LF
+ *   | GC_LV
+ *   | GC_LVT
+ *   | GC_Prepend
+ *   | GC_Regional_Indicator
+ *   | GC_SpacingMark
+ *   | GC_T
+ *   | GC_V
+ *   | GC_ZWJ
+ * )} GraphemeCategoryNum
+ */
+/**
+ * @typedef {(
+ *   | 'Any'
+ *   | 'CR'
+ *   | 'Control'
+ *   | 'Extend'
+ *   | 'Extended_Pictographic'
+ *   | 'L'
+ *   | 'LF'
+ *   | 'LV'
+ *   | 'LVT'
+ *   | 'Prepend'
+ *   | 'Regional_Indicator'
+ *   | 'SpacingMark'
+ *   | 'T'
+ *   | 'V'
+ *   | 'ZWJ'
+ * )} GraphemeCategoryKey
+ */
+/**
+ * Grapheme category enum
+ *
+ * Note:
+ *   The object isn't actually frozen
+ *   because using `Object.freeze` increases 800 bytes on Brotli compression.
+ *
+ * @type {Readonly<Record<GraphemeCategoryKey, GraphemeCategoryNum>>}
+ */
+export const GraphemeCategory: Readonly<Record<GraphemeCategoryKey, GraphemeCategoryNum>>;
+/**
+ * Encoded `Grapheme_Cluster_Break` property ranges.
+ *
+ * Internal categories (never exposed through the public API):
+ *   - 15: InCB_Consonant
+ *
+ * @type {UnicodeDataEncoding}
+ */
+export const grapheme_data: UnicodeDataEncoding;
+/**
+ * Category of each range in {@link grapheme_data}, base36 digit per range.
+ */
+export const grapheme_cats: "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";
+/**
+ * UAX #29 pair-wise rules for extended grapheme cluster boundaries,
+ * evaluated into a 16x16 mask table indexed by `catBefore << 4 | catAfter`.
+ *
+ * A boundary exists iff the mask shares no bit with the runtime's packed sequence state.
+ *
+ * - 0: boundary (GB999 and friends), no state bit can match
+ * - 1: no boundary; bit 0 is always set in the state
+ * - 2: GB12/GB13, no boundary iff odd run of RI precedes
+ * - 4: GB11, no boundary iff the ZWJ was preceded by ExtPic Extend*
+ * - 8: GB9c, no boundary iff InCB Consonant [Extend Linker]* Linker [Extend Linker]* precedes
+ *
+ * See `buildGraphemePairTable` in scripts/unicode.js for the rules.
+ */
+export const grapheme_pairs: "0001000000010010000000100000000000000000000000000001000000010018000100000001001000010101100101100000000000000000000100000001111000010000000110101001110111111111000100000021001000010000000100100001000000011010000100000001111000014000000100180001000000010010";
+export type UnicodeDataEncoding = import("./core.js").UnicodeDataEncoding;
+export type GC_Any = 0;
+export type GC_CR = 1;
+export type GC_Control = 2;
+export type GC_Extend = 3;
+export type GC_Extended_Pictographic = 4;
+export type GC_L = 5;
+export type GC_LF = 6;
+export type GC_LV = 7;
+export type GC_LVT = 8;
+export type GC_Prepend = 9;
+export type GC_Regional_Indicator = 10;
+export type GC_SpacingMark = 11;
+export type GC_T = 12;
+export type GC_V = 13;
+export type GC_ZWJ = 14;
+export type GraphemeCategoryNum = (GC_Any | GC_CR | GC_Control | GC_Extend | GC_Extended_Pictographic | GC_L | GC_LF | GC_LV | GC_LVT | GC_Prepend | GC_Regional_Indicator | GC_SpacingMark | GC_T | GC_V | GC_ZWJ);
+export type GraphemeCategoryKey = ("Any" | "CR" | "Control" | "Extend" | "Extended_Pictographic" | "L" | "LF" | "LV" | "LVT" | "Prepend" | "Regional_Indicator" | "SpacingMark" | "T" | "V" | "ZWJ");

package/node_modules/unicode-segmenter/_grapheme_data.js

@@ -0,0 +1,123 @@
+// The following code was generated by "scripts/unicode.js",
+// DO NOT EDIT DIRECTLY.
+//
+// @ts-check
+
+/**
+ * @typedef {import('./core.js').UnicodeDataEncoding} UnicodeDataEncoding
+ */
+
+/**
+ * @typedef {0} GC_Any
+ * @typedef {1} GC_CR
+ * @typedef {2} GC_Control
+ * @typedef {3} GC_Extend
+ * @typedef {4} GC_Extended_Pictographic
+ * @typedef {5} GC_L
+ * @typedef {6} GC_LF
+ * @typedef {7} GC_LV
+ * @typedef {8} GC_LVT
+ * @typedef {9} GC_Prepend
+ * @typedef {10} GC_Regional_Indicator
+ * @typedef {11} GC_SpacingMark
+ * @typedef {12} GC_T
+ * @typedef {13} GC_V
+ * @typedef {14} GC_ZWJ
+ * @typedef {(
+ *   | GC_Any
+ *   | GC_CR
+ *   | GC_Control
+ *   | GC_Extend
+ *   | GC_Extended_Pictographic
+ *   | GC_L
+ *   | GC_LF
+ *   | GC_LV
+ *   | GC_LVT
+ *   | GC_Prepend
+ *   | GC_Regional_Indicator
+ *   | GC_SpacingMark
+ *   | GC_T
+ *   | GC_V
+ *   | GC_ZWJ
+ * )} GraphemeCategoryNum
+ */
+
+/**
+ * @typedef {(
+ *   | 'Any'
+ *   | 'CR'
+ *   | 'Control'
+ *   | 'Extend'
+ *   | 'Extended_Pictographic'
+ *   | 'L'
+ *   | 'LF'
+ *   | 'LV'
+ *   | 'LVT'
+ *   | 'Prepend'
+ *   | 'Regional_Indicator'
+ *   | 'SpacingMark'
+ *   | 'T'
+ *   | 'V'
+ *   | 'ZWJ'
+ * )} GraphemeCategoryKey
+ */
+
+/**
+ * Grapheme category enum
+ *
+ * Note:
+ *   The object isn't actually frozen
+ *   because using `Object.freeze` increases 800 bytes on Brotli compression.
+ *
+ * @type {Readonly<Record<GraphemeCategoryKey, GraphemeCategoryNum>>}
+ */
+export const GraphemeCategory = {
+  Any: 0,
+  CR: 1,
+  Control: 2,
+  Extend: 3,
+  Extended_Pictographic: 4,
+  L: 5,
+  LF: 6,
+  LV: 7,
+  LVT: 8,
+  Prepend: 9,
+  Regional_Indicator: 10,
+  SpacingMark: 11,
+  T: 12,
+  V: 13,
+  ZWJ: 14,
+};
+
+/**
+ * Encoded `Grapheme_Cluster_Break` property ranges.
+ *
+ * Internal categories (never exposed through the public API):
+ *   - 15: InCB_Consonant
+ *
+ * @type {UnicodeDataEncoding}
+ */
+export const grapheme_data = /** @type {UnicodeDataEncoding} */ (
+  '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'
+);
+
+/**
+ * Category of each range in {@link grapheme_data}, base36 digit per range.
+ */
+export const grapheme_cats = '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';
+
+/**
+ * UAX #29 pair-wise rules for extended grapheme cluster boundaries,
+ * evaluated into a 16x16 mask table indexed by `catBefore << 4 | catAfter`.
+ *
+ * A boundary exists iff the mask shares no bit with the runtime's packed sequence state.
+ *
+ * - 0: boundary (GB999 and friends), no state bit can match
+ * - 1: no boundary; bit 0 is always set in the state
+ * - 2: GB12/GB13, no boundary iff odd run of RI precedes
+ * - 4: GB11, no boundary iff the ZWJ was preceded by ExtPic Extend*
+ * - 8: GB9c, no boundary iff InCB Consonant [Extend Linker]* Linker [Extend Linker]* precedes
+ *
+ * See `buildGraphemePairTable` in scripts/unicode.js for the rules.
+ */
+export const grapheme_pairs = '0001000000010010000000100000000000000000000000000001000000010018000100000001001000010101100101100000000000000000000100000001111000010000000110101001110111111111000100000021001000010000000100100001000000011010000100000001111000014000000100180001000000010010';

package/node_modules/unicode-segmenter/core.cjs

@@ -0,0 +1,135 @@
+"use strict";
+
+exports.decodeUnicodeData = decodeUnicodeData;
+exports.decodeUnicodeFlatData = decodeUnicodeFlatData;
+exports.findUnicodeRangeCategory = findUnicodeRangeCategory;
+// @ts-check
+
+/**
+ * @template {number} [T=number]
+ * @typedef {[from: number, to: number, category: T]} CategorizedUnicodeRange
+ */
+
+/**
+ * @typedef {CategorizedUnicodeRange<0>} UnicodeRange
+ */
+
+/**
+ * @typedef {string & { __tag: 'UnicodeDataEncoding' }} UnicodeDataEncoding
+ *
+ * Encoding for array of {@link UnicodeRange}, as a separator-free
+ * sequence of variable-length quantities in the base36 alphabet;
+ * each character carries 4 payload bits and a continuation bit.
+ *
+ * Each {@link UnicodeRange} packed as a pair of quantities:
+ *
+ * gap      = from - (previous to) - 1
+ * padding  = to - from
+ *
+ * Notes:
+ * - Ranges are sorted and never overlap, so the delta-encoded gap is
+ *   always non-negative, and mostly a single character.
+ * - The max value of a category is 23; https://www.unicode.org/reports/tr29/tr29-45.html#Table_Word_Break_Property_Values
+ * - The longest range is 42,720; CJK UNIFIED IDEOGRAPH-20000..CJK UNIFIED IDEOGRAPH-2A6DF
+ *
+ * See `scripts/lib/encoding.js` for the encoder.
+ */
+
+/**
+ * Decode {@link UnicodeDataEncoding} and hand every range to `emit`.
+ *
+ * Callers build their own index from the callback, so decoding never materializes an intermediate array of ranges;
+ * on the grapheme table that is ~800 short-lived tuples per module load.
+ *
+ * @param {UnicodeDataEncoding} data
+ * @param {string} cats base36 category digit per range, or `''` for none
+ * @param {(from: number, to: number, cat: number) => void} emit
+ * @return {void}
+ */
+function decodeUnicodeData(data, cats, emit) {
+  let len = data.length,
+    i = 0,
+    j = 0,
+    p = -1;
+  let read = () => {
+    let n = 0,
+      shift = 0,
+      d;
+    do {
+      d = parseInt(data[i++], 36);
+      n += (d & 15) << shift;
+      shift += 4;
+    } while (d & 16);
+    return n;
+  };
+  while (i < len) {
+    let from = p + 1 + read();
+    emit(from, p = from + read(), cats ? parseInt(cats[j++], 36) : 0);
+  }
+}
+
+/**
+ * @template {object} Ext
+ * @typedef {{
+ *   segment: string,
+ *   index: number,
+ *   input: string,
+ * } & Ext} SegmentOutput
+ */
+
+/**
+ * @template {object} T
+ * @typedef {IterableIterator<SegmentOutput<T>>} Segmenter
+ */
+
+/**
+ * @typedef {[starts: Uint32Array, ends: Uint32Array]} UnicodeRangeTable
+ *
+ * Flat, binary-searchable range table. `starts[i]` holds the first
+ * codepoint of the i-th range, and `ends[i]` packs its last codepoint
+ * and category as `end << 5 | category`.
+ *
+ * The 5-bit category fits any break property; the largest category
+ * value is 23.
+ */
+
+/**
+ * Category of the range containing the given codepoint.
+ *
+ * @param {number} cp
+ * @param {Uint32Array} starts sorted, non-overlapping range starts
+ * @param {Uint32Array} ends packed `end << 5 | category`, parallel to `starts`
+ * @return {number} category of the containing range, or 0 if no match
+ */
+function findUnicodeRangeCategory(cp, starts, ends) {
+  let lo = 0,
+    hi = starts.length - 1;
+  while (lo <= hi) {
+    let mid = lo + hi >>> 1,
+      end = ends[mid];
+    if (cp < starts[mid]) hi = mid - 1;else if (cp > end >>> 5) lo = mid + 1;else return end & 31;
+  }
+  return 0;
+}
+
+/**
+ * Decode {@link UnicodeDataEncoding} into a membership test table for
+ * {@link findUnicodeRangeCategory}; every range takes category 1, so a
+ * match is truthy and a miss is 0.
+ *
+ * @param {UnicodeDataEncoding} data
+ * @return {UnicodeRangeTable}
+ */
+function decodeUnicodeFlatData(data) {
+  // Every range takes at least two characters,
+  // so half the encoded length is a safe upper bound for the range count.
+  let cap = data.length >> 1,
+    starts = new Uint32Array(cap),
+    ends = new Uint32Array(cap),
+    n = 0;
+  decodeUnicodeData(data, '', (from, to) => {
+    starts[n] = from;
+    ends[n++] = to << 5 | 1;
+  });
+  return [starts.slice(0, n), ends.slice(0, n)];
+}

package/node_modules/unicode-segmenter/core.d.cts

@@ -0,0 +1 @@
+export * from "./core.js";

package/node_modules/unicode-segmenter/core.d.ts

@@ -0,0 +1,118 @@
+/**
+ * @template {number} [T=number]
+ * @typedef {[from: number, to: number, category: T]} CategorizedUnicodeRange
+ */
+/**
+ * @typedef {CategorizedUnicodeRange<0>} UnicodeRange
+ */
+/**
+ * @typedef {string & { __tag: 'UnicodeDataEncoding' }} UnicodeDataEncoding
+ *
+ * Encoding for array of {@link UnicodeRange}, as a separator-free
+ * sequence of variable-length quantities in the base36 alphabet;
+ * each character carries 4 payload bits and a continuation bit.
+ *
+ * Each {@link UnicodeRange} packed as a pair of quantities:
+ *
+ * gap      = from - (previous to) - 1
+ * padding  = to - from
+ *
+ * Notes:
+ * - Ranges are sorted and never overlap, so the delta-encoded gap is
+ *   always non-negative, and mostly a single character.
+ * - The max value of a category is 23; https://www.unicode.org/reports/tr29/tr29-45.html#Table_Word_Break_Property_Values
+ * - The longest range is 42,720; CJK UNIFIED IDEOGRAPH-20000..CJK UNIFIED IDEOGRAPH-2A6DF
+ *
+ * See `scripts/lib/encoding.js` for the encoder.
+ */
+/**
+ * Decode {@link UnicodeDataEncoding} and hand every range to `emit`.
+ *
+ * Callers build their own index from the callback, so decoding never materializes an intermediate array of ranges;
+ * on the grapheme table that is ~800 short-lived tuples per module load.
+ *
+ * @param {UnicodeDataEncoding} data
+ * @param {string} cats base36 category digit per range, or `''` for none
+ * @param {(from: number, to: number, cat: number) => void} emit
+ * @return {void}
+ */
+export function decodeUnicodeData(data: UnicodeDataEncoding, cats: string, emit: (from: number, to: number, cat: number) => void): void;
+/**
+ * @template {object} Ext
+ * @typedef {{
+ *   segment: string,
+ *   index: number,
+ *   input: string,
+ * } & Ext} SegmentOutput
+ */
+/**
+ * @template {object} T
+ * @typedef {IterableIterator<SegmentOutput<T>>} Segmenter
+ */
+/**
+ * @typedef {[starts: Uint32Array, ends: Uint32Array]} UnicodeRangeTable
+ *
+ * Flat, binary-searchable range table. `starts[i]` holds the first
+ * codepoint of the i-th range, and `ends[i]` packs its last codepoint
+ * and category as `end << 5 | category`.
+ *
+ * The 5-bit category fits any break property; the largest category
+ * value is 23.
+ */
+/**
+ * Category of the range containing the given codepoint.
+ *
+ * @param {number} cp
+ * @param {Uint32Array} starts sorted, non-overlapping range starts
+ * @param {Uint32Array} ends packed `end << 5 | category`, parallel to `starts`
+ * @return {number} category of the containing range, or 0 if no match
+ */
+export function findUnicodeRangeCategory(cp: number, starts: Uint32Array, ends: Uint32Array): number;
+/**
+ * Decode {@link UnicodeDataEncoding} into a membership test table for
+ * {@link findUnicodeRangeCategory}; every range takes category 1, so a
+ * match is truthy and a miss is 0.
+ *
+ * @param {UnicodeDataEncoding} data
+ * @return {UnicodeRangeTable}
+ */
+export function decodeUnicodeFlatData(data: UnicodeDataEncoding): UnicodeRangeTable;
+export type CategorizedUnicodeRange<T extends number = number> = [from: number, to: number, category: T];
+export type UnicodeRange = CategorizedUnicodeRange<0>;
+/**
+ *
+ * Encoding for array of {@link UnicodeRange}, as a separator-free
+ * sequence of variable-length quantities in the base36 alphabet;
+ * each character carries 4 payload bits and a continuation bit.
+ *
+ * Each {@link UnicodeRange} packed as a pair of quantities:
+ *
+ * gap      = from - (previous to) - 1
+ * padding  = to - from
+ *
+ * Notes:
+ * - Ranges are sorted and never overlap, so the delta-encoded gap is
+ *   always non-negative, and mostly a single character.
+ * - The max value of a category is 23; https://www.unicode.org/reports/tr29/tr29-45.html#Table_Word_Break_Property_Values
+ * - The longest range is 42,720; CJK UNIFIED IDEOGRAPH-20000..CJK UNIFIED IDEOGRAPH-2A6DF
+ *
+ * See `scripts/lib/encoding.js` for the encoder.
+ */
+export type UnicodeDataEncoding = string & {
+    __tag: "UnicodeDataEncoding";
+};
+export type SegmentOutput<Ext extends object> = {
+    segment: string;
+    index: number;
+    input: string;
+} & Ext;
+export type Segmenter<T extends object> = IterableIterator<SegmentOutput<T>>;
+/**
+ * Flat, binary-searchable range table. `starts[i]` holds the first
+ * codepoint of the i-th range, and `ends[i]` packs its last codepoint
+ * and category as `end << 5 | category`.
+ *
+ * The 5-bit category fits any break property; the largest category
+ * value is 23.
+ */
+export type UnicodeRangeTable = [starts: Uint32Array, ends: Uint32Array];

package/node_modules/unicode-segmenter/core.js

@@ -0,0 +1,129 @@
+// @ts-check
+
+/**
+ * @template {number} [T=number]
+ * @typedef {[from: number, to: number, category: T]} CategorizedUnicodeRange
+ */
+
+/**
+ * @typedef {CategorizedUnicodeRange<0>} UnicodeRange
+ */
+
+/**
+ * @typedef {string & { __tag: 'UnicodeDataEncoding' }} UnicodeDataEncoding
+ *
+ * Encoding for array of {@link UnicodeRange}, as a separator-free
+ * sequence of variable-length quantities in the base36 alphabet;
+ * each character carries 4 payload bits and a continuation bit.
+ *
+ * Each {@link UnicodeRange} packed as a pair of quantities:
+ *
+ * gap      = from - (previous to) - 1
+ * padding  = to - from
+ *
+ * Notes:
+ * - Ranges are sorted and never overlap, so the delta-encoded gap is
+ *   always non-negative, and mostly a single character.
+ * - The max value of a category is 23; https://www.unicode.org/reports/tr29/tr29-45.html#Table_Word_Break_Property_Values
+ * - The longest range is 42,720; CJK UNIFIED IDEOGRAPH-20000..CJK UNIFIED IDEOGRAPH-2A6DF
+ *
+ * See `scripts/lib/encoding.js` for the encoder.
+ */
+
+/**
+ * Decode {@link UnicodeDataEncoding} and hand every range to `emit`.
+ *
+ * Callers build their own index from the callback, so decoding never materializes an intermediate array of ranges;
+ * on the grapheme table that is ~800 short-lived tuples per module load.
+ *
+ * @param {UnicodeDataEncoding} data
+ * @param {string} cats base36 category digit per range, or `''` for none
+ * @param {(from: number, to: number, cat: number) => void} emit
+ * @return {void}
+ */
+export function decodeUnicodeData(data, cats, emit) {
+  let len = data.length
+    , i = 0
+    , j = 0
+    , p = -1;
+  let read = () => {
+    let n = 0, shift = 0, d;
+    do {
+      d = parseInt(data[i++], 36);
+      n += (d & 15) << shift;
+      shift += 4;
+    } while (d & 16);
+    return n;
+  };
+  while (i < len) {
+    let from = p + 1 + read();
+    emit(from, p = from + read(), cats ? parseInt(cats[j++], 36) : 0);
+  }
+}
+
+/**
+ * @template {object} Ext
+ * @typedef {{
+ *   segment: string,
+ *   index: number,
+ *   input: string,
+ * } & Ext} SegmentOutput
+ */
+
+/**
+ * @template {object} T
+ * @typedef {IterableIterator<SegmentOutput<T>>} Segmenter
+ */
+
+/**
+ * @typedef {[starts: Uint32Array, ends: Uint32Array]} UnicodeRangeTable
+ *
+ * Flat, binary-searchable range table. `starts[i]` holds the first
+ * codepoint of the i-th range, and `ends[i]` packs its last codepoint
+ * and category as `end << 5 | category`.
+ *
+ * The 5-bit category fits any break property; the largest category
+ * value is 23.
+ */
+
+/**
+ * Category of the range containing the given codepoint.
+ *
+ * @param {number} cp
+ * @param {Uint32Array} starts sorted, non-overlapping range starts
+ * @param {Uint32Array} ends packed `end << 5 | category`, parallel to `starts`
+ * @return {number} category of the containing range, or 0 if no match
+ */
+export function findUnicodeRangeCategory(cp, starts, ends) {
+  let lo = 0, hi = starts.length - 1;
+  while (lo <= hi) {
+    let mid = lo + hi >>> 1
+      , end = ends[mid];
+    if (cp < starts[mid]) hi = mid - 1;
+    else if (cp > end >>> 5) lo = mid + 1;
+    else return end & 31;
+  }
+  return 0;
+}
+
+/**
+ * Decode {@link UnicodeDataEncoding} into a membership test table for
+ * {@link findUnicodeRangeCategory}; every range takes category 1, so a
+ * match is truthy and a miss is 0.
+ *
+ * @param {UnicodeDataEncoding} data
+ * @return {UnicodeRangeTable}
+ */
+export function decodeUnicodeFlatData(data) {
+  // Every range takes at least two characters,
+  // so half the encoded length is a safe upper bound for the range count.
+  let cap = data.length >> 1
+    , starts = new Uint32Array(cap)
+    , ends = new Uint32Array(cap)
+    , n = 0;
+  decodeUnicodeData(data, '', (from, to) => {
+    starts[n] = from;
+    ends[n++] = to << 5 | 1;
+  });
+  return [starts.slice(0, n), ends.slice(0, n)];
+}

package/node_modules/unicode-segmenter/emoji.cjs

@@ -0,0 +1,43 @@
+"use strict";
+
+exports.isEmoji = isEmoji;
+exports.isEmojiPresentation = isEmojiPresentation;
+exports.isExtendedPictographic = isExtendedPictographic;
+var _core = require("./core.cjs");
+var _emoji_data = require("./_emoji_data.cjs");
+// @ts-check
+
+const [EMOJI_PRESENTATION_S, EMOJI_PRESENTATION_E] = _emoji_data.emoji_presentation_table;
+const [EXTENDED_PICTOGRAPHIC_S, EXTENDED_PICTOGRAPHIC_E] = _emoji_data.extended_pictographic_table;
+
+/**
+ * An alias to {@link isExtendedPictographic}
+ *
+ * @deprecated in favor of {@link isExtendedPictographic}, will be removed in v1.
+ *
+ * @param {number} cp
+ * @return boolean
+ */
+function isEmoji(cp) {
+  return isExtendedPictographic(cp);
+}
+
+/**
+ * Check if the given code point is included in Unicode \\p{Emoji_Presentation} script property
+ *
+ * @param {number} cp
+ * @return boolean
+ */
+function isEmojiPresentation(cp) {
+  return (0, _core.findUnicodeRangeCategory)(cp, EMOJI_PRESENTATION_S, EMOJI_PRESENTATION_E) !== 0;
+}
+
+/**
+ * Check if the given code point is included in Unicode \\p{Extended_Pictographic} script property
+ *
+ * @param {number} cp
+ * @return boolean
+ */
+function isExtendedPictographic(cp) {
+  return (0, _core.findUnicodeRangeCategory)(cp, EXTENDED_PICTOGRAPHIC_S, EXTENDED_PICTOGRAPHIC_E) !== 0;
+}

package/node_modules/unicode-segmenter/emoji.d.cts

@@ -0,0 +1 @@
+export * from "./emoji.js";

package/node_modules/unicode-segmenter/emoji.d.ts

@@ -0,0 +1,23 @@
+/**
+ * An alias to {@link isExtendedPictographic}
+ *
+ * @deprecated in favor of {@link isExtendedPictographic}, will be removed in v1.
+ *
+ * @param {number} cp
+ * @return boolean
+ */
+export function isEmoji(cp: number): boolean;
+/**
+ * Check if the given code point is included in Unicode \\p{Emoji_Presentation} script property
+ *
+ * @param {number} cp
+ * @return boolean
+ */
+export function isEmojiPresentation(cp: number): boolean;
+/**
+ * Check if the given code point is included in Unicode \\p{Extended_Pictographic} script property
+ *
+ * @param {number} cp
+ * @return boolean
+ */
+export function isExtendedPictographic(cp: number): boolean;