rose-ash/hosts/ocaml/bootstrap.py

#!/usr/bin/env python3
"""
Bootstrap compiler: SX spec -> OCaml.

Loads the SX-to-OCaml transpiler (transpiler.sx), feeds it the spec files,
and produces sx_ref.ml — the transpiled evaluator as native OCaml.

Usage:
    python3 hosts/ocaml/bootstrap.py --output hosts/ocaml/lib/sx_ref.ml
"""
from __future__ import annotations

import os
import sys

_HERE = os.path.dirname(os.path.abspath(__file__))
_PROJECT = os.path.abspath(os.path.join(_HERE, "..", ".."))
sys.path.insert(0, _PROJECT)

from shared.sx.parser import parse_all
from shared.sx.types import Symbol


def extract_defines(source: str) -> list[tuple[str, list]]:
    """Parse .sx source, return list of (name, define-expr) for top-level defines."""
    exprs = parse_all(source)
    defines = []
    for expr in exprs:
        if isinstance(expr, list) and expr and isinstance(expr[0], Symbol):
            if expr[0].name == "define":
                name = expr[1].name if isinstance(expr[1], Symbol) else str(expr[1])
                defines.append((name, expr))
    return defines


# OCaml preamble — opens and runtime helpers
PREAMBLE = """\
(* sx_ref.ml — Auto-generated from SX spec by hosts/ocaml/bootstrap.py *)
(* Do not edit — regenerate with: python3 hosts/ocaml/bootstrap.py *)

[@@@warning "-26-27"]

open Sx_types
open Sx_runtime

(* Trampoline — forward ref, resolved after eval_expr is defined. *)
let trampoline_fn : (value -> value) ref = ref (fun v -> v)
let trampoline v = !trampoline_fn v



(* === Mutable state for strict mode === *)
(* These are defined as top-level refs because the transpiler cannot handle
   global set! mutation (it creates local refs that shadow the global). *)
let _strict_ref = ref (Bool false)
let _prim_param_types_ref = ref Nil

"""


# OCaml fixups — wire up trampoline + iterative CEK run
FIXUPS = """\

(* Wire up trampoline to resolve thunks via the CEK machine *)
let () = trampoline_fn := (fun v ->
  match v with
  | Thunk (expr, env) -> eval_expr expr (Env env)
  | _ -> v)

(* Override recursive cek_run with iterative loop *)
let cek_run_iterative state =
  let s = ref state in
  while not (match cek_terminal_p !s with Bool true -> true | _ -> false) do
    s := cek_step !s
  done;
  cek_value !s

"""


def compile_spec_to_ml(spec_dir: str | None = None) -> str:
    """Compile the SX spec to OCaml source."""
    from shared.sx.ref.sx_ref import eval_expr, trampoline, make_env, sx_parse

    if spec_dir is None:
        spec_dir = os.path.join(_PROJECT, "spec")

    # Load the transpiler
    env = make_env()
    transpiler_path = os.path.join(_HERE, "transpiler.sx")
    with open(transpiler_path) as f:
        transpiler_src = f.read()
    for expr in sx_parse(transpiler_src):
        trampoline(eval_expr(expr, env))

    # Spec files to transpile (in dependency order)
    # stdlib.sx functions are already registered as OCaml primitives —
    # only the evaluator needs transpilation.
    sx_files = [
        ("evaluator.sx", "evaluator (frames + eval + CEK)"),
    ]

    parts = [PREAMBLE]

    for filename, label in sx_files:
        filepath = os.path.join(spec_dir, filename)
        if not os.path.exists(filepath):
            print(f"Warning: {filepath} not found, skipping", file=sys.stderr)
            continue

        with open(filepath) as f:
            src = f.read()
        defines = extract_defines(src)

        # Skip defines provided by preamble, fixups, or already-registered primitives
        # Skip: preamble-provided, math primitives, and stdlib functions
        # that use loop/named-let (transpiler can't handle those yet)
        skip = {"trampoline", "ceil", "floor", "round", "abs", "min", "max",
                "debug-log", "debug_log", "range", "chunk-every", "zip-pairs",
                "string-contains?", "starts-with?", "ends-with?",
                "string-replace", "trim", "split", "index-of",
                "pad-left", "pad-right", "char-at", "substring"}
        defines = [(n, e) for n, e in defines if n not in skip]

        # Deduplicate — keep last definition for each name (CEK overrides tree-walk)
        seen = {}
        for i, (n, e) in enumerate(defines):
            seen[n] = i
        defines = [(n, e) for i, (n, e) in enumerate(defines) if seen[n] == i]

        # Build the defines list for the transpiler
        defines_list = [[name, expr] for name, expr in defines]
        env["_defines"] = defines_list

        # Pass known define names so the transpiler can distinguish
        # static (OCaml fn) calls from dynamic (SX value) calls
        env["_known_defines"] = [name for name, _ in defines]

        # Call ml-translate-file — emits as single let rec block
        translate_expr = sx_parse("(ml-translate-file _defines)")[0]
        result = trampoline(eval_expr(translate_expr, env))

        parts.append(f"\n(* === Transpiled from {label} === *)\n")
        parts.append(result)

    parts.append(FIXUPS)
    output = "\n".join(parts)

    # Post-process: fix mutable globals that the transpiler can't handle.
    # The transpiler emits local refs for set! targets within functions,
    # but top-level globals (*strict*, *prim-param-types*) need to use
    # the pre-declared refs from the preamble.
    import re

    # Fix *strict*: use _strict_ref instead of immutable let rec binding
    output = re.sub(
        r'and _strict_ =\n  \(Bool false\)',
        'and _strict_ = !_strict_ref',
        output,
    )
    # Fix set-strict!: use _strict_ref instead of local ref
    output = re.sub(
        r'and set_strict_b val\' =\n  let _strict_ = ref Nil in \(_strict_ := val\'; Nil\)',
        "and set_strict_b val' =\n  _strict_ref := val'; Nil",
        output,
    )
    # Fix *prim-param-types*: use _prim_param_types_ref
    output = re.sub(
        r'and _prim_param_types_ =\n  Nil',
        'and _prim_param_types_ = !_prim_param_types_ref',
        output,
    )
    # Fix set-prim-param-types!: use _prim_param_types_ref
    output = re.sub(
        r'and set_prim_param_types_b types =\n  let _prim_param_types_ = ref Nil in \(_prim_param_types_ := types; Nil\)',
        "and set_prim_param_types_b types =\n  _prim_param_types_ref := types; Nil",
        output,
    )

    # Fix all runtime reads of _strict_ and _prim_param_types_ to deref
    # the mutable refs instead of using the stale let-rec bindings.
    # This is needed because let-rec value bindings capture initial values.
    # Use regex with word boundary to avoid replacing _strict_ref with
    # !_strict_refref.
    def fix_mutable_reads(text):
        lines = text.split('\n')
        fixed = []
        for line in lines:
            # Skip the definition lines
            stripped = line.strip()
            if stripped.startswith('and _strict_ =') or stripped.startswith('and _prim_param_types_ ='):
                fixed.append(line)
                continue
            # Replace _strict_ as a standalone identifier only (not inside
            # other names like set_strict_b). Match when preceded by space,
            # paren, or start-of-line, and followed by space, paren, or ;.
            line = re.sub(r'(?<=[ (])_strict_(?=[ );])', '!_strict_ref', line)
            line = re.sub(r'(?<=[ (])_prim_param_types_(?=[ );])', '!_prim_param_types_ref', line)
            fixed.append(line)
        return '\n'.join(fixed)
    output = fix_mutable_reads(output)

    return output


def main():
    import argparse
    parser = argparse.ArgumentParser(description="Bootstrap SX spec -> OCaml")
    parser.add_argument(
        "--output", "-o",
        default=None,
        help="Output file (default: stdout)",
    )
    args = parser.parse_args()

    result = compile_spec_to_ml()

    if args.output:
        with open(args.output, "w") as f:
            f.write(result)
        size = os.path.getsize(args.output)
        print(f"Wrote {args.output} ({size} bytes)", file=sys.stderr)
    else:
        print(result)


if __name__ == "__main__":
    main()