Example programs
Once you've done the tutorial, these fuller programs are a good next step. They all live in the repository's examples/ directory, and each one's README explains how to run it, what it depends on and what its limits are.
Start with small local tools
- Task Tracker: a command-line task list that keeps its data in a local JSON file. It shows file I/O, a typed data model and reading JSON fields with
@std/json_codec, with no network access. - json_select: a command-line tool for picking fields out of JSON, split across several files, using the
sprig-clilibrary for options. - tasks: a todo CLI with subcommands over a JSON file (
sprig-cli,@std/json_codec). - csv: writes a UTF-8 CSV file and reads it back with
sprig-csv, with commas, a line break and non-ASCII text inside cells, and reads its columns by header. The library ships in the SDK from v0.9.0-beta.1. - config_summary: reads a small JSON configuration file and prints a summary in a fixed format, with clear errors for bad input.
- blog: a static blog generator that turns posts with front matter into HTML pages and an index, with exit codes for bad input (
@std/files,@std/text,@std/lists). - agent_tools: three small tools written in Sprig that read the compiler's JSON output to query Java and Sprig APIs, summarize errors and compare APIs.
Applications that use Java libraries
- application_foundation: HTTP, JSON encoding and decoding, UTC time and file handling.
- sqlite and sqlite_migrations: JDBC through Maven, with transactions, persistence and database migrations.
- parallel_words: counts words in parallel with
@std/concurrent, using a scope, tasks, a pool, a channel, a counter and a lock once each; see Concurrency. - crawler: a concurrent crawler over local pages with scopes, tasks, a channel, a counter and a lock (
@std/concurrent). - ledger: a compact bookkeeping HTTP backend whose data survives a restart. See Web and SQLite.
- todo: a todo list stored in SQLite behind a small HTTP API (
sprig-sqlite,sprig-web); its test drives the live server over HTTP. - mini_web: typed routes, JSON and OpenAPI documentation.
Third-party Java libraries are added with ordinary Maven coordinates and pinned by the lock file; see projects and JVM interop. The libraries Sprig maintains itself are in the libraries/ directory.
Bigger programs
- repository_audit: walks a code repository and writes a JSON report.
- maven_slug: locks a Maven library, queries its API and reuses the cache offline.
- source_analyzer: a source analysis tool written in Sprig that walks a syntax tree and lists functions, variables and references. It covers part of the language only and isn't a full compiler API.
- test_runner: all kinds of tests in an ordinary Sprig project, including runtime tests, table tests, temporary files, subprocesses and tests that are expected to fail compilation.
Fabric mods is a separate case study: game logic in a Minecraft mod written in Sprig, with Gradle and Loom handling the build and only a thin layer of Java.
These examples cover the scenarios that have been tested. They don't mean every library API or production workload is supported.
