Research-Stack/4-Infrastructure/shim/build_corpus278.py

#!/usr/bin/env python3
# /// script
# requires-python = ">=3.10"
# dependencies = []
# ///
"""
Build Semantics/RRC/Corpus278.lean from rrc_equation_classifier_receipt.json.

Python's role: read raw fields, map to stable IDs, derive generator tokens.
Lean's role: alignment gate, receipt stamping, all output decisions.
"""
from __future__ import annotations
import hashlib, json, re, sys
from pathlib import Path

ROOT = Path("/home/allaun/Research Stack")
RECEIPT_JSON = ROOT / "archive/experimental-shim-probes/rrc_equation_classifier_receipt.json"
OUT_LEAN = ROOT / "0-Core-Formalism/lean/Semantics/Semantics/RRC/Corpus278.lean"

# ── shape mapping (classifier JSON → Lean RRCShape constructor) ──────────────
SHAPE_MAP = {
    "CognitiveLoadField":                   ".cognitiveLoadField",
    "SignalShapedRouteCompiler":             ".signalShapedRouteCompiler",
    "ProjectableGeometryTopology":          ".projectableGeometryTopology",
    "CadForceProbeReceipt":                 ".cadForceProbeReceipt",
    "LogogramProjection":                   ".logogramProjection",
    "HoldForUnlawfulOrUnderspecifiedShape": ".holdForUnlawfulOrUnderspecifiedShape",
}

# ── template_key logic (rrc_kind → template) ─────────────────────────────────
def template_key(rrc_kind: str, status: str) -> str:
    if status == "HOLD":
        return "hold"
    kind_map = {
        "cognitive_field_receipt":  "definition",
        "compression_route_prior":  "master_equation",
        "geometry_topology_receipt":"definition",
        "cad_force_receipt":        "gate",
        "logogram_projection":      "receipt",
        "negative_control":         "hold",
    }
    return kind_map.get(rrc_kind, "definition")

# ── operator token derivation (route_hint + rrc_kind + equation text) ─────────
def operator_tokens(er: dict) -> list[str]:
    tokens = []
    rh = (er.get("route_hint_non_authoritative") or "").strip()
    rk = (er.get("rrc_kind") or "").strip()
    if rh and rh != "unclassified_equation":
        tokens.append(rh)
    if rk:
        tokens.append(rk)
    # Extract simple operator keywords from equation text
    eq_text = (er.get("equation") or "").lower()
    for op in ["exp(", "log(", "max(", "min(", "sum(", "integral", "derivative",
               "laplacian", "nabla", "div(", "curl(", "sigmoid", "softmax",
               "tanh(", "relu(", "norm(", "dot(", "cross("]:
        if op in eq_text:
            tokens.append(op.rstrip("("))
    return list(dict.fromkeys(tokens))  # deduplicate, preserve order

# ── string escaping for Lean ──────────────────────────────────────────────────
def lean_str(s: str) -> str:
    s = s.replace("\\", "\\\\").replace('"', '\\"')
    return f'"{s}"'

def lean_opt(s: str | None) -> str:
    if s is None:
        return "none"
    return f"some {lean_str(s)}"

def lean_str_list(xs: list[str]) -> str:
    return "[" + ", ".join(lean_str(x) for x in xs) + "]"

# ── stable equation ID ────────────────────────────────────────────────────────
def stable_id(raw_id: str) -> str:
    h = hashlib.sha256(raw_id.encode()).hexdigest()[:16]
    return f"rrc_eq_{h}"

# ── main ──────────────────────────────────────────────────────────────────────
def main() -> None:
    d = json.loads(RECEIPT_JSON.read_text())
    eqs = d["compiled_equations"]
    print(f"Loaded {len(eqs)} equations", file=sys.stderr)

    lines: list[str] = []
    lines.append("-- Semantics.RRC.Corpus278 — AUTO-GENERATED by build_corpus278.py")
    lines.append("-- DO NOT EDIT BY HAND.  Regenerate with:")
    lines.append("--   python3 4-Infrastructure/shim/build_corpus278.py")
    lines.append("--")
    lines.append("-- Python role: raw feature extraction only.")
    lines.append("-- Lean role: alignment gate, receipt stamping, all output decisions.")
    lines.append("--")
    lines.append("-- Source: archive/experimental-shim-probes/rrc_equation_classifier_receipt.json")
    lines.append(f"-- Equation count: {len(eqs)}")
    lines.append("")
    lines.append("import Semantics.RRC.Emit")
    lines.append("")
    lines.append("namespace Semantics.RRC.Corpus278")
    lines.append("")
    lines.append("open Semantics.RRC.Emit")
    lines.append("open Semantics.RRCLogogramProjection")
    lines.append("open Semantics.ReceiptCore")
    lines.append("")
    lines.append("/-- Full 278-equation corpus from rrc_equation_classifier_receipt.json.")
    lines.append("    Each row carries raw features only; the alignment gate in")
    lines.append("    Semantics.RRC.Emit.emitCorpus makes all admissibility decisions. -/")
    lines.append("def corpus278 : List FixtureRow := [")

    row_strs: list[str] = []
    for eq in eqs:
        er  = eq["equation_record"]
        ir  = eq["invariant_receipt"]
        tw  = eq["type_witness"]

        raw_id   = er["equation_id"]
        eq_id    = stable_id(raw_id)
        name     = er["name"]
        shape_str = ir["shape"]
        lean_shape = SHAPE_MAP.get(shape_str, ".holdForUnlawfulOrUnderspecifiedShape")
        status_str = ir["status"]
        lean_status = ".candidate" if status_str == "CANDIDATE" else ".hold"
        rrc_kind = er.get("rrc_kind", "")
        weak_cnt = len(tw.get("missing_or_weak_axes") or [])

        # Generator fields
        op_tokens    = operator_tokens(er)
        inv_declared = (er.get("domain_type") or "unknown").strip() or "unknown"
        bound_conds  = (er.get("bind_class") or "unknown").strip() or "unknown"
        t_key        = template_key(rrc_kind, status_str)
        t_params     = f"route={er.get('route_hint_non_authoritative','unclassified_equation') or 'unclassified_equation'};shape={shape_str}"

        row = (
            f"  {{ equationId          := {lean_str(eq_id)}\n"
            f"    name                := {lean_str(name)}\n"
            f"    shape               := {lean_shape}\n"
            f"    status              := {lean_status}\n"
            f"    rrcKind             := {lean_str(rrc_kind)}\n"
            f"    weakAxesCnt         := {weak_cnt}\n"
            f"    pistProxyLabel      := none\n"
            f"    pistExactLabel      := none\n"
            f"    operatorTokens      := {lean_str_list(op_tokens)}\n"
            f"    invariantsDeclared  := {lean_str(inv_declared)}\n"
            f"    boundaryConds       := {lean_str(bound_conds)}\n"
            f"    templateKey         := {lean_str(t_key)}\n"
            f"    templateParams      := {lean_str(t_params)} }}"
        )
        row_strs.append(row)

    lines.append(",\n".join(row_strs))
    lines.append("]")
    lines.append("")
    lines.append("end Semantics.RRC.Corpus278")
    lines.append("")

    OUT_LEAN.write_text("\n".join(lines))
    print(f"Wrote {OUT_LEAN}", file=sys.stderr)
    print(f"Total rows: {len(row_strs)}", file=sys.stderr)

if __name__ == "__main__":
    main()