llama-compiler

Llama is a relatively simple language that combines the functional and imperative programming models. It is based on a subset of OCaml, with which it shares many similarities. Its main features in brief are the following:

• Powerful type system with type inference.
• Basic data types for integers, characters, boolean values and real numbers.
• Pointers and tables of one or more dimensions. The memory locations where references and array elements point are mutable and assignable.
• Programmer-defined data types, possibly recursive as well.
• High-order functions, but without partial evaluation. Passing parameters by value.
• Function library.

Requirements

Software	Version
make	4.3
flex	2.6.4
bison	3.8.2
g++	11.2.0
llvm-12	12.0.1

Usage

./llamac [-O] <file.lla>

./llamac [options]

Options

Flag	Description
-O	Optimization flag.
-f	Input from stdin, final code in stdout.
-i	Input from stdin, intermediate code in stdout.
-p	Input from stdin, AST in stdout.

Examples

./llamac -O ./programs/hello.lla

./llamac -O -f

./llamac -i

Note: You first need to compile the compiler.

Linking

gcc <file.s> ./lib/lib.a -o <file.out> -lm -no-pie

Note: You first need to compile the library (i.e., run make inside the lib dir).