UniqueNamesGenerator.jl

Generate unique, memorable, human-readable names from configurable word-list dictionaries. Name generation (word selection) is cleanly separated from name rendering (formatting into a string), so the same generated name can be displayed in multiple formats without regenerating.

Quick Start

using UniqueNamesGenerator

# 1. Generate a name and display it
name = generate()                # → UniqueName(["brave", "elegant", "dolphin"])
println(name)                    # "Brave Elegant Dolphin"
string(name)                     # "Brave Elegant Dolphin"

# 2. Inspect the raw words
name.words                       # ["brave", "elegant", "dolphin"]

# 3. Render the same name in several formats (keyword form)
name = generate([["swift"], ["falcon"]])
render(name)                                           # "Swift Falcon"
render(name; separator = "-", style = :lowercase)      # "swift-falcon"
render(name; separator = "_", style = :uppercase)      # "SWIFT_FALCON"

# 4. Render via RenderOptions and the shipped presets
render(name, TITLE)              # "Swift Falcon"
render(name, SLUG)               # "swift-falcon"
render(name, SCREAMING_SNAKE)    # "SWIFT_FALCON"

# 5. Batch generation
generate([ADJECTIVES, ANIMALS], 3; unique = true)

# 6. Exclusion set
exclude = Set([UniqueName(["swift", "falcon"])])
generate([["swift", "brave"], ["falcon"]], exclude)    # UniqueName(["brave", "falcon"])

# 7. Deterministic RNG
using Random
generate([ADJECTIVES, ANIMALS]; rng = MersenneTwister(42))

Generate vs. render

• generate picks one word from each input dictionary and returns a UniqueName holding the raw Vector{String} of selected words. No formatting happens here.
• render turns a UniqueName into a String. It accepts either keyword arguments (separator, style) or a RenderOptions value. Both forms share a single rendering kernel and produce byte-identical output for the same (separator, style) pair.
• TITLE, SLUG, and SCREAMING_SNAKE are exported const RenderOptions presets for the most common formats.

Built-in Dictionaries

The word lists are maintained in a separate data repository: unique-names-data.

Decoupling data from code allows the dictionaries to be updated, enriched, or reused by other projects (in any language) without releasing a new version of the Julia package. The data is fetched at build time via the Julia Artifacts system and cached locally.

Custom Dictionaries

Use load_dictionary to load a CSV file (expects word,category header):

my_dict = load_dictionary("path/to/words.csv")
generate([ADJECTIVES, my_dict])