Lookup

animated-fluent-emojis/lookup has no React and shares the manifest with Emoji, so it is cheap to add next to it. Every function loads the manifest and resolves to undefined or an empty array, never rejects, when it cannot:

import {
  extractEmojis,
  findEmojiByUnicode,
  searchEmojis,
} from 'animated-fluent-emojis/lookup'

await findEmojiByUnicode('👍🏽') // { id: 'yes', skinTone: 'medium' }
await extractEmojis('Hi 👋 there') // [{ id, text, index, length }]
await searchEmojis('party', { limit: 5 }) // [{ id }]

• findEmojiByUnicode(text) resolves one emoji and maps a single skin tone modifier to skinTone; mixed tones resolve to the base emoji. Symbols such as © or ™ need the emoji variation selector (U+FE0F) to match, while ZWJ sequences match even when the variation selector is missing (minimally qualified).
• When several catalog entries share a glyph, lookup returns the canonical emoji: the id prefixed with the glyph’s code points, otherwise a reviewed override, otherwise the first entry in catalog order. For example, ❤️ resolves to the heart rather than a variant that reuses the glyph. With a skin tone, it falls back to a sibling entry that has tones.
• extractEmojis(text) finds every catalog emoji in a text, keeping ZWJ sequences whole, with its offset and length. Without Intl.Segmenter it falls back to a code point grouper, and neither function ever rejects.
• searchEmojis(query, { limit }) matches descriptions, ignoring case; limit defaults to 20; a limit that is not a positive number means no limit, except 0, which returns nothing.