Hello, Crowfile!

Single files are fine for scripts; projects get a directory. crow new scaffolds one:

crow new hello_crow
cd hello_crow

Look at what it made — two files:

hello_crow/
  crowfile
  main.co

The crowfile is the project manifest. It is itself a Crow file — one assignment for now:

name = 'hello_crow'

Later you'll add keys to it: db = '...' so the checker can read your database schema (chapter 11), jobs = map[...] for batch tasks, deps for packages. main.co is the entry point, and it already runs:

crow run          # compile and run main.co
crow run -i       # the same program, interpreted — instant start
crow build        # compile only; the binary lands in the project directory

crow run writes the generated Rust into .crow_build/ inside the project. You never need to look at it, and it's safe to delete — the next run recreates it. The first compile takes a few seconds; after that only what changed is rebuilt.

Two more commands you'll use in every chapter from here on:

crow check        # verify the project: your tests plus everything automatic
crow fmt          # format every .co file in the project

For a web project, crow new myapp --web adds a views/ directory wired into the server (chapter 12); --db and --ml add database and training scaffolds. Every scaffolded file is working code.

You now have everything the rest of this book needs. Next: build a real program before learning any more theory.