Search code, repositories, users, issues, pull requests...

Also comes bundled with Emacs. Just C-h i (or M-x info) and open Emacs Lisp Intro.

A bit scholarly. Not the easiest read.

Emacs In A Box - Elisp Programming

Very complete and gentle introduction.

Written by Caio Rordrigues (@caiorss).

(Almost) All You Need to Know About Variables

A must read!

elisp-guide

Focused on introducing general concepts and terminology.

Emergency Elisp

Focused around data structures. More like a cheat sheet.

xahlee.info - Practical Emacs Lisp

read online, accompagnying youtube live streams

One of the first resources on the web introducing Elisp in a concise and comprehensible way.

The website used to be called ergoemacs.

Still updated regularly.

Please note that a few more complex examples follow an older coding style: big monolithic functions instead of multiple small functional ones.

ElispCheatSheet

Focused around data structures.

Author’s Common Lisp background can be felt.

Advanced

Emacs Lisp Reference Manual

Also comes bundled with Emacs. Just C-h i (or M-x info) and open Elisp.

Comprehensive guide of core Elisp APIs.

Generally well written but needs you to understand its logic and be familiar with Emacs terminology.

The Emacs Lisp Style Guide

Provides solid guidelines on the dos and don’ts for quality Elisp code.

The Emacs Package Developer’s Handbook

@alphapapa’s organized notes about Elisp development.

List bunch of tools and libraries for helping package developers.

Also provides various snippets and best practices.

Lots of good insights.

nullprogram.com

Chris Wellons (@skeeto)’s blog, especially the posts tagged #elisp.

A few highlights:

Emacs Lisp Programming Thoughts

Pretty old but has very good insights, guidelines and real-world examples.

Byte compiler section is outdated but general coding style recommendations and sections about macros and profiling are still relevant to this day.

Cookbooks

elisp-demos

Very good and beginner-friendly.

Almost complete list of single-liner examples of standard function. Also packs examples for dash and s.

Emacs In A Box - Elisp Snippets

Really nice selection of snippets with real-world use-cases.

EmacsWiki’s Cookbook

Community-driven snippets, beginner-friendly.

@alphapapa’s unpackaged.el

Real-world selection of snippets, not beginner-friendly.

On Package Authoring

The Emacs Lisp Style Guide applies all the more in this context.

Article: Take Your Emacs to the Next Level by Writing Custom Packages

Real world experience of a user writing and submitting his first package.

MELPA recommandations

There’s a high chance that you’ll be uploading your package on MELPA.

They have clear recommandations.

Don’t worry, for your first submissions, they will be very comprehensive and will help you fixing what’s wrong.

Contributing to Emacs

The CONTRIBUTE file is the official document describing the process. Additional development tips and coding conventions can be found in the Elisp Manual.

M-x view-emacs-todo shows a lists of TODO items you might want to work on. You can also browse the bug archive using M-x debbugs-gnu using the debbugs package.

Contributing to Emacs gives some helpful background information and overview about the contribution workflow for newcomers.

Development Tools

By default, Emacs is already pretty well set up for Elisp development.

But some features can be hard to learn and some stuff can be improved with additinal packages.

See also those talks John Wiegley gave about his setup for Elisp development:

Interactive Development & Debugging

Emacs is built with interactive development in mind.

You could spend days developing Elisp code without ever having to restart Emacs.

Standard commands used are:

eval-last-sexp (C-x C-e)
eval-defun (C-M-x)
eval-buffer
eval-region

The *scratch* buffer also provides a temporary zone to try and test ideas. In it can be used eval-print-last-sexp (C-j) which acts like eval-last-sexp but also prints the result after the s-exp in the buffer.

eval-expression (M-:) allows quickly evaluating a s-exp from anywhere by entering it in the minibuffer.

For logging, function (message "<text>") allows printing into the *Messages* buffer.

For debugging, the most basic command is toggle-debug-on-error to get a stacktrace.

IELM

Stands for Inferior Emacs Lisp Mode.

Provides a REPL for evaluating Elisp code.

edebug

source, doc

edebug is the interactive Elisp debugger.

The documentation is a bit rough to get started. I recommend reading this series of posts:

You can also read the dedicated chapter in book An Introduction to Programming in Emacs Lisp.

trace

Provides a simple facility to output a trace of function calls into a buffer (*trace-buffer*).

Please note that this trace is very basic and has no performance data. For more advanced tracing see Profiling.

Tracing is switched on / off with commands trace-function, untrace-function and untrace-all.

macrostep

Interactive macro expander.

Expand nested macros one by one.

Way better than using default macroexpand.

eval-expr

Provides eval-expr, an enhanced eval-expression command.

Some highlights:

automatic display of output in temp buffer if too big
allows pretty printing of results (with pp)
invalid s-expr don’t have to be retyped on 2nd try

eval-sexp-fu

Visual improvment.

Flashes the sexps during the evaluation.

eros

Show evaluation results inline.

Documentation & Introspection

To get the documentation of a symbol, you could use one of the built-in:

describe-symbol: get documentation of symbol
describe-function: get documentation of function
describe-variable: get documentation of variable
describe-key: get documentation of command associated with keybinding

These would spawn a *Help* buffer. Hence documentation in Emacs is often referred to as the help. For searching through symbols Emacs also comes with various apropos-* commands which populate a buffer with your search results.

Some packages improve on these.

helpful

Provides more contextual information.

helpful command	default command	comment
`helpful-at-point`	`describe-symbol`
`helpful-callable`	no equivalent	like `helpful-function` but also works on macros and special forms
`helpful-function`	`describe-function`
`helpful-macro`	no equivalent
`helpful-variable`	`describe-variable`
`helpful-key`	`describe-key`

elisp-demos

Provides usage examples in the *Help* buffer.

which-key

which-key is like an always-on describe-key.

It displays automatically all the possible keybindings following a key prefix.

suggest

Discover elisp functions by specifying input and the desired output.

Code Editing

Jump to definition

To jump to the definition of a symbol Emacs provides xref-find-definitions. In practice it works with nicely with functions but is kind of hit-or-miss with variables.

In addition, the following more specialised functions exist:

find-function: go to the definition of function
find-variable: go to the definition of variable
find-library: go to the definition of feature (i.e. module, package)

Better options exists so that you don’t have to remember all of these.

Honorable mention: elisp-slime-nav, that can be seen as an ancestor to elisp-def. If you want to jump to symbols in files which aren’t loaded in your Emacs you can fallback to the more general dumb-jump package.

elisp-def

Provides elisp-def that allows jumping to the definition of function / variable / feature.

Like a better xref-find-definitions.

Is able to distinguish between functions / variables / features depending on the context.

Also handles macros, functions defined through macros and let-bound variables.

Search & Replace

In Emacs regular expressions can make use of syntax information provided by the major-mode. This means that to some extend semantic searches are possible with isearch and occur. To search and jump to toplevel definitions of a buffer you can use the built-in imenu. The imenu-anywhere package allows to extend the scope to buffers of the same project or mode.

elisp-refs

Semantic code search for Elisp which parses the code instead of doing dump text searches.

el-search

Lets you execute search and replace operations on symbolic expressions. For example you can search for things like defvars which don’t specify an init value using the pattern `(defvar ,_)`.

Editing

Honorable mentions:

YASnippet: generic (not Elisp-specific) powerful abreviation-based snippet expander. Even though it could be used in the place of speed-of-thought-lisp, it is less context-aware and requires a specific key combination to trigger. It offers other features, though, and can be used complementarily.

speed-of-thought-lisp

Allows writting Elisp blazingly fast with the use of context-aware abbreviations triggered after <SPACE> keypress.

elisp-docstring

Enriched syntax highlighting for docstring contents. Together with string-edit you can edit docstrings in a temporary buffer and get automated special character escaping.

Editing S-exps

Elisp is a Lisp and Lisps are written using a structure of nested lists called S-expressions.

Mastering how to navigate and manipulate this structure with ease is essential.

By default Emacs doesn’t offer much apart from forward-list / backward-list and forward-sexp / backward-sexp.

Luckily, powerful minor-modes are available to give you the power you deserve.

lispy

source & doc, cheat sheet, video demos

Easiest to learn yet most powerful solution in that list.

Like paxedit, commands are context-aware.

The killing feature is that shortcuts are single characters and not key combinations.

The trick is that commands get triggered only when the point is at a delimiter (e.g. a parenthesis) or the region is active.

Provides a powerful command combination system and refactoring commands.

symex

Builds on paredit, lispy and evil-cleverparens to provide a consistent vim-style modal way of editing trees.

Offers a hydra interface to aid learning, and an evil state (also works in vanilla emacs).

paxedit

Heavily inspired by paredit.

The major difference with the latter is that commands are context-aware, they behave differently depending on what the cursor is hovering on.

The direct consequence is that fewer commands / shortcuts needs to be learned to perform the same amount of things.

paredit

tutorial, source

The first powerful S-exp editing mode for Emacs.

Learning curve is a bit steep.

Still actively maintained and very popular.

Refactoring

elisp-depmap

Aids the refactoring process by presenting a graphical visualization of project dependencies.

emacs-refactor

Contains various refactoring commands for Elisp.

Formatting

aggressive-indent-mode

Auto-indents code as you type.

elfmt

doc, blog post on nullprogram.com

Focuses on placement of lists and tries to break lines at fill-column

Testing

For simulating interactive user input, consider using libraries such as with-simulated-input (launch commands) and dokey (simulated keyboard shortcut presses). To test behavior and interactive usage in a clean and temporary environment emacs-sandbox.sh is useful.

ERT

Stands for “Emacs Lisp Regression Testing”.

Featureful and easy to use.

Suitable for unit tests with mocking.

Buttercup

Behavior-Driven Emacs Lisp Testing

Especially suitable for integration tests.

Allows defining test suites (i.e. goups of related tests) with a shared context (through set-up and tear-down phases).

Also provides mocking capabilities.

For a complete example of integration w/ undercover and coverage:

Sacha Chua - Coverage reporting in Emacs with Buttercup, Undercover, Coverage, and a Makefile

director

Program sequence of user interactions. Useful for end-to-end testing.

Inspired by Selenium Webdriver.

undercover

Track test coverage and integrate w/ coverage reporting solutions.

For in-Emacs reporting, use the coverage package.

Profiling

Emacs provides 2 Elips profilers:

profiler.el: profile a whole call stack, easier to use
elp.el: profile only selected functions

Both are briefly mentioned in the profiling section of the Emacs Lisp Reference Manual.

Either one is of a great help to debug slow Elisp code, most noticeable during user interactions (Emacs seems to freeze).

profiler.el is easily toggled using profiler-start, profiler-stop. To obtain a result report call profiler-report.

elp.el can target individual functions with elp-instrument-function or a whole package with elp-instrument-package.

To profile individual forms Emacs also comes with the benchmark library. The Emacs Package Developer’s Handbook has a whole section dedicated to this with thorough examples and helper macros.

etrace

Wrapper around elp.el outputting a report in the Chromium Catapult Trace Event Format.

This allows opening them in external applications to explore them as flame graphs.

Validation

Emacs provides various functions to validate an Elisp file / project:

byte-compile-file: validate the file compiles cleanly
checkdoc: validate the documentation
check-declare-file / check-declare-directory: validate the declaration of symbols
package-lint-current-buffer: validate format for submitting as a package

It’s tedious to run manually each and every of those commands. Thankfully projects aim at making this process easier.

For maximum efficiency, they can be integrated into a CI chain (GitHub actions or Travis).

melpazoid

In addition to standard validation, it adds a license checker and some additional checks.

Created by MELPA member @riscy to validate submissions.

Does not run tests.

Provides recipes for integration with GitHub actions or Travis.

makem.sh

Very straightforward way to validate an Emacs package folder / repository.

Provides a makefile with different targets to run.

Implemented in bash with a makefile wrapper.

Performs linting (make lint), tests (make test) or everything (make all).

In addition to standard checks, also validates indentation and optionally elsa checks.

Supports both ERT and buttercup tests.

One drawback is that this makem.sh sources have to be dropped in each of your package source repository.

Provides recipes for integration with GitHub actions.

makel

Provides a makefile with different targets to run.

Implemented completely as a makefile.

Requires a bit of configuration for each package.

One drawback is that this makel sources have to be dropped in each of your package source repository.

No CI integration recipes.

elisp-check

A zero config github action to validate packages.

auto-compile

Compiles current file on save and display compile errors/warnings in the mode-line.

elisp-lint

Performs standard validation of specified file. Also checks for indentation.

No CI integration recipes.

elsa

Static Elisp code analyzer providing helpful hints.

Can be launched directly from makem.sh.

package-lint

Lints Elisp files for requirements of packages. Can be integrated with flycheck (a general linter framework) by installing flyspell-package.

Building

Those tools, in addition to what those in the Validation section provide, are full-fledged build-definition tools, allowing to make complex CI/CD chains.

They require a fair amount of configuration and are not for the faint of heart.

They only seem necessary when building larger packages with exotic dependencies.

Eldev

Stands for “Elisp Development Tool”.

Certainly the most modern of the lot.

100% written in Elisp.

One small drawback is that it does not run in a dedicated isolated Emacs process.

cask

Pretty advanced and hard to get into.

Implemented in python.

Runs in a dedicated isolated Emacs process

emake

The most simple to use from this list.

Implemented in Elisp with a makefile wrapper.

Easier to integrate with CI tools such as Travis.

Elisp Repo Kit

elisp-repo-kit

Create new fully-featured Github repositories with minimal effort, including doc generation and CI.

Integrates common development & distribution workflows such as testing & package reloading.

Designed to work out of the box but also appropriate for complex 3rd party and non-Elisp module integrations.

Syntax Highlighting & Visual Enhancements

Several packages provide visual improvements and extend default syntax highlighting (font locking in Emacs lingo).

All those listed bellow are complementary.

Honorable mentions:

highlight-defined which is superseded by lisp-extra-font-lock functionalities

Not Elisp-specific but commonly used in the context of Elisp development:

traditionally, form feed characters (^L) are used in Elisp source as a section delimiters. Either page-break-lines or form-feed can be used to display them as intended.
for those that can barely stand parentheses, paren-face can be used to dim them in Lisp-based modes
for those that love parentheses, rainbow-delimiters allows displaying them in different colors depending on their nesting depth

lisp-extra-font-lock

Various additional syntax highlightings.

Killer feature is having different faces for special vars (global) VS normal ones (local).

highlight-function-calls

Make functions calls stand out with a specific face.

cl-lib-highlight

Provides additional / alternative font-locking for cl-lib symbols, to make them stand out in your code.

Also highlights deprecated cl symbols with a different face. Useful when reading legacy code.

easy-escape

Make regular expression strings more readable.

nameless

Hide prefix in symbols of a package.

Pretty Printing

pp

Standard Emacs pretty-printing util.

ppp

Advanced pretty-printing utils.

Libraries

Traditionally, it was recommended to not use external libs/dependencies and prefer using standard APIs bundled with Emacs.

These recommendation are still mostly valid but predated the advent of package.el.

Some external libs are now considered “standard”, as lots of popular packages use them and they can outperform standard implementations while still being simpler to use (e.g. dash).

Some libraries might be listed several times, as they fit in several categories (e.g. subr-x, dash).

Core / General Purpose

cl-lib

Lib extending Elisp with functionalities inherited from Common Lisp. Replaces the deprecated cl package which did not use name prefixes. To help with updating the code from cl to cl-lib there is cl-libify.

Just do a (require 'cl-lib) to use it.

subr-x

Intended as an extension to subr.el, the core library of basic functions written in Elisp.

Provides:

threading macros (a la Clojure, thread-first and thread-last)
additional binding helpers (if-let, if-let*, when-let, when-let* and and-let*) - hash-table manipulation helper (hash-table-empty-p, hash-table-keys and hash-table-values)
string manipulation helper (string-empty-p, string-blank-p, string-join, string-trim, string-trim-left, string-trim-right, string-remove-prefix and string-remove-suffix)
region manipulation helpers (replace-region-contents)

dash

Even though this lib revolves primarily around list manipulation, it also offers for general purpose utils.

Those are:

They all seem to be heavily inspired by Clojure.

el-patch

source and doc

More perene advices, get notified when they break.

anaphora

Allows the definition of anaphoric functions (as can be found in Common Lisp, Clojure…).

with-simulated-input

Simulate user interactions (i.e. launch commands).

Mostly usefull for writing tests.

dokey

Trigger keyboard events.

contract

Provides data structure defintions as contracts (essentially interface description).

Port of Racket’s contract to Elisp.

Akin to Clojure’s spec.

signal

Reimplementation of hooks, with more advanced features.

weak-ref

Allows creating weak reference to vars. Weak reference offer better performance but can be garbage collected.

predd

Provides Clojure-style multimethods (multiple dispatch over an ad hoc type hierarchy).

cats

Provices Haskell-inspired Category Theory abstractions.

fn

Provides macros for a more concise lambda syntax, a la Clojure.

Data Structures

Strings

subr-x

Provide the following helpers: string-empty-p, string-blank-p, string-join, string-trim, string-trim-left, string-trim-right, string-remove-prefix and string-remove-suffix.

s

Advanced yet easy to use string manipulation helpers.

rx

tutorial, source

Macro for helping writing Elisp regexp.

xr

Convert regexp to their more human-readable rx macro form.

Also provides regexp linting, detecting mistakes and bad practices.

Relies on its own internal parser.

Sequences

seq

source, doc

(since version 25)

dash

Advanced yet easy to use list manipulation helpers. Lots of them also have alternative anaphoric forms.

stream

Allows defining streams of data as data sequences. Compatible w/ seq.el.

trie

Provides APIs for building and manipulating tries, sequence-like data structures where both storage and retrieval are space- and time-efficient.

Stored elements must be ordered sequences, i.e. strings (most common use-case), lists or vectors.

lister

Provides lister-mode, major mode for building and manipulating list-based user-interfaces.

Inspired by tablist (for tables) and hierarchy (for trees).

Maps

(Hash)maps are a special type of sequences that allow representing a list of key / value pairs. In other languages they can also be called associative arrays or dictionaries.

In Elisp, a map can be represented as:

an alist (association list, preserving element order)
a plist (property list, more human-readable)
an hash-table

data structure	human-readability	insert speed	lookup speed	ordered?
alist	meh	fastest	slower as data grows	yes
plist	very good	ok	fast	no
hash-table	ok	ok	very fast	no

The official doc also has a nice section comparing plists and alists.

tl;dr:

planning on doing lots of inserts and a few lookups (or mostly on recent elements), use an alist
planning on having a big number of elements and lookup speed is critical, use an hash-map
every other case, use a plist

Older Emacs packages tend to rely mostly on alists, sometimes for no good reason.

Each data structure has its own APIs to get/insert/update.

Thankfully, some libraries provide an abstraction layer that allows having a single API for multiple data structures.

I would recommend sticking with the default map.el library, unless you really enjoy the Clojure syntax in which case a.el is also a nice choice. If you know for sure that you want to stick with an alist or a hash-table, asoc.el and ht are high quality choices.

map

(since version 25)

supports: alists, plists and hash-tables.

Shared API for all 3 Elisp map objects (+ arrays).

No documentation other than what is inlined in source.

asoc

supports: only alists.

Nice set of additional APIs for alists.

ht

supports: only hash-tables, but allow converting from/to alists and plists.

Nice set of additional APIs for hash-tables.

a

supports: alists and hash-tables.

Shared API for alists and hash-tables. Like map.el, but in a more “Clojurey” syntax.

kv

support: mostly alists, but allow converting from/to alists and plists.

dict-tree

Provides APIs for building and manipulating Dictionary trees, hybrid between tries and hash tables.

Think about it as a more storage-efficient hash tables.

Custom Types & OOP

Can be done natively using records, additional custom user-defined types.

cl-lib (defstruct API)

API documentation

One part of cl-lib is APIs to define and manipulate C-like data structures, strongly typed.

Provides the cl-defstruct macro.

Built on top of the native records system.

See also this blog post from @skeeto: Options for Structured Data in Emacs Lisp

EIEIO

doc

Stands for Enhanced Implementation of Emacs Interpreted Objects.

Brings an OOP layer to Elisp, based upon the Common Lisp Object System (CLOS).

Provides the defclass macro.

Built on top of the native records system.

Date & Time

ts

Advanced yet easy to use datetime / timestamp library.

datetime

Library for parsing, formatting, matching and recoding timestamps and date-time format strings.

datetime-format

Provides datetime-format, inspired by PHP’s Datetime::format method.

Tables

tabulated-list

Library for defining, manipulating and displaying tables.

tablist

Extension to tabulated-list, adding possibility to mark and filter items.

navigel

Facilitate the creation of tabulated-list-based UIs.

Also relies on tablist.

gridlock

Provides gridlock-mode, major mode for building and manipulating spreadsheet-based user-interfaces

Also provides gridlock-csv-mode and gridlock-fix-mode minor modes, backporting the API to CSV and FIX files.

cell

Provides cell-mode, major mode for building and manipulating spreadsheet-based user-interfaces.

ctable

Library for defining, manipulating and displaying tables.

Queues

queue

Provides FIFO / FILO queue APIs.

fifo-class

An EIEIO abstract class class to provide FIFO methods to slots.

Rings

Even though ring is the standard implementation, some core libs use their own internal implementation (e.g. the kill-ring).

ring

Provides APIs to create and manipulate a ring data structure.

Used by: ERC

dynaring

Similar to ring, but w/ a dynamic size.

Trees

Escaped and nested S-exps is the most straightforward way to encode a tree in (E)lisp.

Some libraries deliver higherèlevel data structure with manipulation functions for improved performance and convenience.

heap

Provides APIs to build and manipulate a ternary (at most 3 children per node) heap (self-sorting tree).

avl-tree

Provides APIs to build and manipulate a self-balancing binary tree.

hierarchy

Allows defining trees as well as building user interfaces displaying them.

treepy

Allows defining and traversing trees.

taxy

Allows defining hierarchical taxonomies, i.e. trees w/ automatic classification based on (nested) rules.

rbit

Self-balancing interval trees.

Implementation of Chris Okasaki’s algorithm from “Red-black trees in a functional setting”, JFP’99.

pair-tree

Visualize and explore nested S-exps as a tree.

Parsers & Parse Trees

Those libraries allow parsing a document in a format / language and converting it to an tree, called an an AST.

parse-it

source & doc, presentation @ EmacsConf21

Regexp-based parser, supporting a bunch of languages.

tree-sitter

doc, source, reddit post, example usage article

Implemented as a module, binding with the tree-sitter parser (written in Rust).

For a concrete use-case, have a look at tree-edit.

For a better sitter grammar for elisp (distinguishing between var, functions and macros), use tree-sitter-elisp.

moldable-emacs

Powerful parser and transformer library, relying on the concept of composable functional molds.

Also support asynchronous processing (relying on async).

tNFA

Provides APIs to build and manipulate NFA (Nondeterministic Finite Automaton), i.e. a state machine / decision tree.

It was built manily with regexp parsing in mind.

parsec

Parsing library in the spirit of Haskell’s parsec.

pl

Parsing library in the spirit of Haskell’s parsec. Somewhat limited.

XML/HTML

dom

DOM manipulation and searching functions.

xml-query

List-based XML selectors. Part of the elfeed package.

Org-mode outlines

org-mode outlines (spec) can be considered both a file format and a tree format.

org-element

source, doc, tutorial on ergoemacs

org-mode’s internal parser, used to convert a text buffer into a tree structure (parse-tree).

org-ml

Functional manipulation of an org parse-tree.

org-ql

Query language (DSL) for parsing, searching and filtering an org outline.

org-ba

More user-friendly APIs for writting code for interacting with org documents.

Faces

Faces are a group of attributes controlling the formatting of text in Emacs.

It’s akin to CSS for HTML or styling ANSI escape sequences for terminal text.

You can read more about faces in the Emacs manual or the Emacs Lisp Reference Manual.

engrave-faces

Convert faces to other formats.

Currently, only LaTeX is supported.

Color Codes

color

color-tools

Color codes manipulation, with support for various color spaces.

yk-color

Color codes manipulation.

Concurrency / Asynchronicity

Concurrency in Elisp / Emacs is a hot topic.

Due to its single-threaded nature, we can’t do parallel processing unless using some dirty tricks (see async).

But that doesn’t prevent us from doing concurrent processing, with say timers.

Fairly recently, Emacs extended this support with generators (since 25.1) and native threads (not what you might be thinking of, since 26.1).

For more info on those subject, read:

emacswiki/No Threading
emacswiki/Concurrent Emacs
blog post from @skeeto: Emacs 26 Brings Generators and Threads

Timers

timer

Default timer lib.

named-timer

Easier to use timer lib.

Promises & Delays

thunk.el

Provides an API for creating and dereferencing / evaluating delays.

promise.el

Reimplementation of the Promises/A+ open standard (originally targeting Javascript).

aio

Mostly an async/await lib but implements its own promise system internally.

Async Elisp function calls

deferred

Not super-actively maintained, but quite stable and featureful.

Achieves concurrency through the use of timers.

Also allows handling async (sub-)processes and HTTP calls with request.el bindings.

concurrent

source, doc

Higher-level wrapper around deferred.

Provides various syntaxes inspired by those of other programming languages, such as:

Clojure / Java / Lua’s coroutines (threads)
Python’s asyncio coroutines (generators)
Clojure’s core.async pipelines (signals / channels).

async

Achieves true parallel processing by spawning a child Emacs sub-process. As such, necessary context needs to be passed w/ async-inject-variables.

Supports defining callbacks.

Offers bindings w/ dired, bytecomp and smtp-mail.

timp

Multithreading through sub-processes with over-the-wire payload capabilities.

Achieves true parallel processing by spawning a child Emacs sub-process for each thread.

aio

Short for async-io.

Allows writing coroutines with the async/await syntax found in Python’s asyncio.

Internal representation relies on its own promise implementation and generators.

async-await

Simple implementation of Async/Await, based on the TypeScript syntax.

Relies on promise.el and generators.

lcr

lcr stands for Lightweight CoRoutines.

Seems to rely on timers.

Async sub-processes

These libs only allow to run asynchronously command processes (as opposed to Elisp function calls).

It can be done in standard with low-level function make-process or derivatives start-process, make-pipe-process and start-process-shell-command.

Some advanced behaviours are hard to program, that’s why wrapper libraries can help you.

Notably:

ensuring the process is launched asynchronously (not blocking Emacs)
configuring callbacks (by binding a sentinel to the process)

deferred

Not super-actively maintained, but featureful.

Also allows handling async Elisp function calls and HTTP calls with request.el bindings.

bpr

Stands for Background Process Runner. Allows running a command process in the background.

Allows advanced callback behaviours.

It relies on start-process-shell-command.

pfuture

Allows running a command process in the background.

Result can be handled either with a future (pfuture-new, pfuture-result) or a callback (pfuture-callback).

It relies on make-pipe-process for the future-based implementation and make-process for the callback one.

Async interpreter commands

Emacs provides a layer on top of make-process for spawning commands from a shell interpreter (i.e. bash or zsh).

These are provided by simple.el (source).

The async version of these command is async-shell-command.

Some advanced behaviours are hard to program, that’s why wrapper libraries can help you.

dtache

Provides dtache-shell-command, a drop-in replacement for async-shell-command that allows command execution to persist even after the Emacs process exits.

Also works on remote hosts.

Relies on dtach to create a persistent session.

friendly-shell-command

friendly-shell-command provides friendly-shell-command-async, a wrapper around async-shell-command with easier access to advanced behaviours thanks to optional keyword arguments.

It notably eases associating a callback to the end of the execution, running on remote hosts and launching with alternative interpreters.

Buffer Manipulation

b

Utility functions for buffer manipulation.

tp

Utilities for helping with manipulating a buffer’s text properties.

m-buffer

List-oriented functions for accessing and manipulating the contents of Emacs buffers.

Filesystem Interactions

f

Modern API for working with files and directories.

Networking

HTTP client

Emacs comes already with an HTTP client, url.el, written in pure Elisp (source), which has a few limitations. It exposes functions url-retrieve-synchronously and url-retrieve (async).

request

If found on the system, uses the cURL binary instead of url.el. Can be customized with request-backend.

Supports a bunch of options exposed clearly with keyword arguments.

Advanced asynchronicity via bindings with deferred.

plz

Uses the cURL binary (curl) instead of url.el.

Supports a bunch of options exposed clearly with keyword arguments.

Supports both synchronous & asynchronous calls, as well as queuing.

When doing async request, the returned handler is a curl process object.

Strong contender for a lightweight alternative to request.

mb-url

Stands for “Multiple Backends for URL package”.

Provides API-compatible replacements to url-retrieve and url-retrieve-synchronously using cURL and HTTPie.

websocket

Websocket (peristent HTTP connection) client for Emacs.

apiwrap

Macros to ease the definition of binding functions to HTTP APIs.

with-proxy

Wrapper for let-binding HTTP proxies.

HTTP server

simple-httpd

A web server written in pure Elisp, serving HTTP.

web-server

A web server written in pure Elisp, serving HTTP APIs bound to Elisp functions (handlers).

RPC server

A Remote Procedure Call server allows Emacs to receive commands from a remote process through a messaging system.

It’s a common strategy of inter-process communication (IPC).

porthole

Start a HTTP-based RPC server under Emacs.

Commands are direct Elisp code to be executed. They can (by default) only be called synchronously.

Messages are encoded in JSON (following the JSON-RPC 2.0 Specification) which makes it support client libraries of almost any language.

Relies on web-server.

EPC

Start a RPC client & server under Emacs.

It implements its own protocol (over TCP) and support both synchronous & asynchronous execution (via bindings with deferred).

Commands are explicitly defined (akin to handlers bound to routes in an HTTP API).

Messages are encoded as Lisp / S-exprs, which makes it more challenging to implement client libraries in non-Lisp languages.

D-Bus

D-Bus is the most popular inter-process communication (IPC) protocol under Linux.

Emacs supports it by default.

dbus

Very boilerplate-y to use.

debase

EIEIO abstractions over dbus for writting easier interaction code.

Database Access

SQL

sql

source, emacswiki

Not a client per se.

Provides sql-mode, a comint-based REPL wrapper supporting various CLI client interpreters.

sqlite

(since version 29)

Recent Emacs now embbeds a native SQLite3 database & the accompagnying client.

emacsql

High-level client to SQLite, PostgreSQL & MySQL.

Queries and schema definitions are written in specific tree-based DSLs, allowing easy programmatic manipulation.

closql

ORM providing mapping between EIEIO and SQLite tables.

Relies on emacsql.

edbi

Client to SQL dialects, using Perl’s Database Interface (DBI) as a connection interface.

In addition to programmatic querying capabilities, provides several major modes for user interactions with database instances.

triples

Abstraction on top of SQL clients (either emacsql or sqlite) to represent & store a graph database.

GUI

Honorable mention: magic-buffer, an executable cookbook on how to use & abuse Emacs’ buffer display engine.

Popups

frog-menu

User selection menu in the form of a popup.

Overlays

ov