#!/usr/bin/env python3
"""Process the large equation database through the PIST pipeline."""
import json, re, sys, hashlib
from pathlib import Path
from collections import Counter


ROOT = Path("/home/allaun/Research Stack/3-Mathematical-Models")


def tokenize(text: str) -> list[str]:
    return re.findall(r"[a-z0-9]+|[+\-*/=^(){}\[\]]", text.lower())


def build_matrix(tokens: list[str]) -> list[list[int]]:
    vocab = sorted(set(tokens))
    if not vocab:
        return [[0]*8]*8
    matrix = [[0]*8 for _ in range(8)]
    for t_i, t_j in zip(tokens, tokens[1:]):
        matrix[vocab.index(t_i) % 8][vocab.index(t_j) % 8] += 1
    return matrix


def spectral_radius(mat: list[list[int]], n_iter: int = 50) -> float:
    n = len(mat)
    v = [1.0] * n
    for _ in range(n_iter):
        w = [sum(mat[i][j] * v[j] for j in range(n)) for i in range(n)]
        norm = max(abs(x) for x in w) if w else 1.0
        if norm == 0: return 0.0
        v = [x / norm for x in w]
    return norm


def classify(lam: float) -> str:
    if lam >= 4.0: return "CognitiveLoadField"
    elif lam >= 1.5: return "SignalShapedRouteCompiler"
    else: return "LogogramProjection"


def extract_equations() -> list[str]:
    """Extract all equation strings from the JSON files."""
    eqs = []
    
    # From unified_9pattern_samples
    try:
        d = json.loads((ROOT / "unified_9pattern_samples.json").read_text())
        for cat, samples in d.items():
            if isinstance(samples, list):
                for s in samples:
                    if isinstance(s, str):
                        eqs.append(s)
                    elif isinstance(s, dict):
                        for v in s.values():
                            if isinstance(v, str) and len(v) > 5:
                                eqs.append(v)
    except: pass
    
    # From unknown_discovery_report
    try:
        d = json.loads((ROOT / "unknown_discovery_report.json").read_text())
        samples = d.get("samples", {})
        for cat, items in samples.items():
            if isinstance(items, list):
                eqs.extend(items)
    except: pass
    
    # From structural_discovery (top structures)
    try:
        d = json.loads((ROOT / "structural_discovery.json").read_text())
        for item in d.get("top_structures", []):
            eq = item.get("equation", {})
            if isinstance(eq, dict):
                s = eq.get("structure", "")
                if s:
                    eqs.append(s)
            elif isinstance(eq, str):
                eqs.append(eq)
    except: pass
    
    return eqs


def main():
    print("Extracting equations...")
    eqs = extract_equations()
    print(f"  Found {len(eqs)} equation strings")
    
    # Deduplicate
    eqs = list(set(eqs))
    print(f"  Unique: {len(eqs)}")
    
    # Process through pipeline
    results = []
    shape_counts = Counter()
    signal_count = 0
    
    for i, eq in enumerate(eqs):
        if i % 1000 == 0 and i > 0:
            print(f"    processed {i}/{len(eqs)}...")
        
        tokens = tokenize(str(eq))
        if not tokens:
            continue
        matrix = build_matrix(tokens)
        lam = spectral_radius(matrix)
        shape = classify(lam)
        mhash = hashlib.sha256(
            json.dumps(matrix, separators=(",", ":")).encode()
        ).hexdigest()[:12]
        
        shape_counts[shape] += 1
        if lam >= 1.5:
            signal_count += 1
        
        results.append({
            "lambda": round(lam, 4),
            "shape": shape,
            "hash": mhash,
            "len": len(eq),
        })
    
    print(f"\n  {'='*50}")
    print(f"  Results: {len(results)} equations processed")
    print(f"  Signal detected: {signal_count}/{len(results)} ({100*signal_count/len(results):.1f}%)")
    print(f"  Shape distribution:")
    for shape, count in shape_counts.most_common():
        print(f"    {shape}: {count} ({100*count/len(results):.1f}%)")
    print(f"  Mean λ: {sum(r['lambda'] for r in results)/len(results):.3f}")
    
    # Summary receipt
    receipt = {
        "schema": "large_equation_pipeline_v1",
        "source": "equations_compressed (decompressed samples)",
        "total_unique": len(eqs),
        "processed": len(results),
        "signals": signal_count,
        "shapes": dict(shape_counts),
        "mean_lambda": round(sum(r['lambda'] for r in results)/len(results), 4) if results else 0,
    }
    out = Path("signatures/large_equation_pipeline.json")
    out.write_text(json.dumps(receipt, indent=2) + "\n")
    print(f"\n  Receipt: {out}")


if __name__ == "__main__":
    main()