wildling

R library and CLI for pattern-based string generation. Zero third-party dependencies (base R only — hand-rolled JSON parser). Requires R 3.5+ with base packages only.

Docs: Website · Sandbox · Syntax · Source

Registry: R-universe

Example

http://${'dev,stage,prod'}\-${'api,web'}#{0-2}.example.${'com,net,org'}/@.html

(The \- is a literal hyphen; bare - would mean “one letter or digit”. @ is one lowercase letter.)

That builds URL-shaped candidates: scheme http://, then environment × service × optional digits × TLD, then a one-letter path page. Three environments, two services, zero–two digits ('', 0–9, 00–99), three TLDs, and a–z → 51948 strings — the kind of list you generate for fuzzing links or probing staging hosts, not type out.

A few of them:

• http://dev-api.example.com/a.html / http://stage-web.example.com/z.html
• http://dev-api0.example.net/a.html / http://prod-web9.example.org/m.html
• http://dev-api00.example.com/a.html / http://prod-web99.example.org/z.html

Named dictionaries (%{'hosts'}) work the same way when the word lists live in files.

Try it in the sandbox, or see pattern syntax for length ranges, dictionaries, and escapes.

Install

install.packages("wildling", repos = "https://dotmonk.r-universe.dev")

Or from GitHub:

remotes::install_github("dotmonk/wildling", subdir = "r", ref = "v2.0.7")

From this repository:

cd r
./build.sh
./bin/wildling "Year 19##"

library(wildling)

w <- create_wildling(c("Year 19##"))
value <- w$`next`()
while (!isFALSE(value)) {
  cat(value, "\n")
  value <- w$`next`()
}

CLI

./bin/wildling "Year 19##"
./bin/wildling --dictionary planets:../dictionaries/planets.txt "%{'planets'}"
./bin/wildling --template ./config.json

Help text and --check output follow docs/cli.md / docs/help.txt.

Build

./build.sh   # Docker (gcc:14-bookworm + r-base-core): copy help.txt + R parse check

Project tests live in ../tests/ and are run with ../test.sh.