pymojis 1.0.1

A lightweight, type-safe Python library for emoji search, transformation, and detection.

pymojis

A small, type-safe Python library for working with emojis: search them by name / code / character, transform text, detect emojis in strings, and convert to HTML.

• Zero runtime dependencies.
• Two install sizes: a tiny default (~230 KB of data) or the full Unicode dataset (~1 MB) via the [full] extra.
• Strict typing — py.typed, full mypy strict compliance.
• Fail-fast API — bad input raises immediately, no silent None returns or warnings.

Install

pip install pymojis              # lightweight: ~1900 emojis
pip install 'pymojis[full]'      # full Unicode coverage: ~3790 emojis

Requires Python 3.12+.

Quick start

from pymojis import PymojisManager

manager = PymojisManager()

# Pick random emojis
print([e.emoji for e in manager.get_random(length=3)])
# → ['😊', '🎉', '🌟']

# Look up by name / code / character
manager.get_by_name("grinning face with smiling eyes")  # → '😄'
manager.get_by_code("1F604")                            # → '😄'
manager.get_by_emoji("😊")                              # → Emoji(...)

# Replace whole-word matches in text
manager.emojifie("I'm sleepy")
# → "I'm 😪"

# Detection
manager.contains_emojis("hello 👋")    # → True
manager.is_emoji("😄")                 # → True
manager.is_emoji("😄😊")               # → False

# HTML hex references
manager.to_html("😵‍💫")
# → "😵‍💫"

To use the full Unicode dataset:

manager = PymojisManager(use_full_dataset=True)
# Requires `pip install 'pymojis[full]'`. Raises DatasetNotFoundError
# with installation instructions otherwise.

API

Method	Returns	Notes
get_random(categories=None, length=1, exclude=None)	list[Emoji]	categories takes precedence over exclude.
get_all_emojis(exclude=None)	list[Emoji]	exclude accepts "complex" or a list of categories.
get_by_code(code)	str | None	Single-codepoint lookup. Case-insensitive.
get_by_name(name)	str | None	Full-name lookup. Case-insensitive.
get_by_category(category)	list[str]	All emojis in a category.
get_by_emoji(emoji)	Emoji | None	Reverse lookup from character to record.
contains_emojis(text)	bool	True if text contains at least one known emoji.
is_emoji(text)	bool	True if text.strip() is a single known emoji.
emojifie(text)	str	Replace whole words with emojis (whose name contains that word).
to_html(emoji)	str	Encode each codepoint as &#xHEX;.

All methods raise TypeError on non-str arguments — no silent None.

Categories

from pymojis import Categories

Categories  # type alias of all valid categories:
#   "Smileys & Emotion", "People & Body", "Animals & Nature",
#   "Food & Drink", "Activities", "Travel & Places", "Objects",
#   "Symbols", "Flags", "Component"

Emoji data model

from pymojis import Emoji

@dataclass-like
class Emoji:
    id: str            # auto-generated UUID
    emoji: str         # the character itself, e.g. "😄"
    name: str          # e.g. "grinning face with smiling eyes"
    code: list[str]    # one or more Unicode codepoints, e.g. ["1F604"]
                       #   (multi-codepoint emojis like ZWJ sequences have len > 1)
    category: str      # e.g. "Smileys & Emotion"
    sub_category: str  # e.g. "face-smiling"

Notes on emojifie

emojifie does whole-word matching. Each word in the input is looked up in an index of emoji-name tokens. Tokens shorter than 3 characters are skipped (otherwise pronouns like "I" and "m" would be replaced by ℹ and Ⓜ). When a word matches multiple emojis, the first one (in dataset order) wins.

manager.emojifie("I'm sleepy")           # → "I'm 😪"  (sleepy → "sleepy face")
manager.emojifie("zzzz xyzzy")           # → "zzzz xyzzy"  (no match, unchanged)

Releases (maintainer notes)

This repo publishes two packages on every git tag: pymojis and pymojis-fulldata. Both must share the same version.

To cut a release:

# 1. Bump version in BOTH pyproject files
$EDITOR pyproject.toml                                  # version = "X.Y.Z"
$EDITOR packages/pymojis-fulldata/pyproject.toml        # version = "X.Y.Z"
# Also bump the [full] extra dependency pin in pyproject.toml.

# 2. Commit, tag, push
git commit -am "Release vX.Y.Z"
git tag vX.Y.Z
git push origin main --tags

CI runs the full make ci pipeline (lint, format, typecheck, tests, build both wheels, twine check) and publishes via PyPI Trusted Publishing (OIDC — no API tokens). One-time setup: register both projects as Trusted Publishers on https://pypi.org pointing at this repo and the workflow .github/workflows/github_ci.yaml.

Development

A Makefile wraps the common workflows:

make install       # uv sync --extra dev --frozen
make lint          # ruff check
make format-check  # ruff format --check
make typecheck     # mypy src
make test          # pytest
make build         # build both wheels into dist/
make twine-check   # twine check dist/*
make ci            # lint + format-check + typecheck + test + build + twine-check
make clean         # remove dist/, build/, *.egg-info

pre-commit is configured: uv run pre-commit install once, and the same checks run on every commit.

Contributing

Issues and PRs welcome. Please run the full check suite above before opening a PR.

License

MIT — see LICENSE.

The bundled emoji metadata derives from the work of Chalda Pnuzig (ISC License, see third_party/).