Input / Output

FlipJump has no read or write instruction. I/O happens at fixed bit-addresses that the runtime watches.

Output

Flipping a specific bit at the I/O opcode address sends one output bit to stdout. The exact address is exposed by stl.startup as the label IO:

ns stl {
    def startup code_start > IO {
        ;code_start
      IO:                       // exposed: this address IS the I/O opcode
        ;0
    }
}

After running stl.startup, the address of IO is known to the rest of the program. To emit a 0 bit, flip the low bit of IO; to emit a 1, flip a different bit. The STL’s stl.output_bit and stl.output_char wrap the bit-flipping pattern.

Bytes are output LSB first by flipping eight consecutive bits in sequence.

Input

The next input bit is automatically loaded by the runtime at the bit-address 3*w + #w (this is dbit + dw from the I/O opcode). Read it by jumping based on the bit value:

.if0 input_bit, branch_for_zero, branch_for_one

The runtime reloads the next bit on each read. End-of-input is signalled by the input slot being permanently 0 (or by the runtime keeping the program looping until interrupted — implementations vary).

Higher-level wrappers

User code rarely touches the IO opcode directly. The STL provides:

• Output: see the per-file page runlib.fj for stl.output_bit, stl.output_char, stl.output.

• Bit-level input/output: bit/input.fj and bit/output.fj.

• Hex-level input/output: hex/input.fj and hex/output.fj.

A complete “Hello, World!” program is just stl.startup_and_init_all + stl.output "Hello\n" + stl.loop. See the Hello World walkthrough for a guided tour.

Reference

The exact I/O addresses (2*w for the opcode, 3*w + #w for the input bit) come from the upstream language design — for more depth see esolangs.org/FlipJump#Input_/_Output.