Abacus

Abacus is a experimental, interpreted functional programming language with a pleasant REPL

Highlights

• Persistent REPL environment with history and an ANSI-colored prompt.
• 64-bit integers, double-precision floats, and booleans with straightforward formatting.
• Arithmetic, comparison, logical, and bitwise operators with predictable precedence.
• Pattern-matched function arms: mix literals and identifiers and the most specific match runs.
• Helpful diagnostics that underline exactly where evaluation failed.

Install

Cargo

cargo install abacus

Try It

Usage

The Abacus calculator REPL and runner

Usage: abc [OPTIONS] [FILE]

Arguments:
  [FILE]  Execute the given Abacus source file

Options:
  -e, --expr <EXPR>  Evaluate a single expression and exit
  -n, --no-color     Disable ANSI color output
  -h, --help         Print help
  -V, --version      Print version

• Start the interactive REPL: abc
• Evaluate a single expression and print the result: abc -e "2 + 2"
• Execute a source file (runs each line in order, sharing the same environment): abc path/to/program.abc
• Disable ANSI colors for scripts or terminals that do not support them: add --no-color (e.g. abc --no-color -e "1/3")
• Color is also controlled by common environment variables: NO_COLOR disables color, CLICOLOR_FORCE=1 forces it, and CLICOLOR=0 disables it unless forced. Color is off when stdout is not a TTY unless forced.
• Equality is type-strict: ints and floats cross-compare, but comparing booleans to numbers raises a type error instead of silently returning false.
• Function calls have a recursion safeguard (default limit 1000, configurable via --recursion-limit or ABACUS_MAX_CALL_DEPTH); unbounded recursion (e.g., inf(n) = n * inf(n)) will eventually raise a recursion-limit diagnostic instead of looping forever.

REPL sessions retain history, coloring, and diagnostic output between entries:

[0x00]: rate = 0.0825
[0x01]: with_tax(amount) = amount * (1 + rate)
[0x02]: with_tax(24.99)
[0x02]: 27.05475
[0x03]: fib(0) = 0
[0x04]: fib(1) = 1
[0x05]: fib(n) = fib(n-1) + fib(n-2)
[0x06]: fib(10)
[0x06]: 55
[0x07]: odd(n) = n % 2 == 1
[0x08]: odd(41)
[0x08]: true

Build From Source

cargo build

Limitations

• Mixed integer and float arithmetic promotes integers to floats. Integers outside the range 2^53 may lose precision in the conversion.

License

This project is licensed under the MIT License.