"""Survey Fractonaut output for interesting/distinctive content."""
import json, sys, re
from collections import Counter

def load(p):
    out = []
    with open(p) as f:
        for line in f:
            line = line.strip()
            if not line: continue
            try: out.append(json.loads(line))
            except: pass
    return out

def text_of(e):
    return (e.get("response") or e.get("observation") or e.get("error") or "").strip()

paths = sys.argv[1:] or [
    "/mnt/d/Resonance_Engine/traj/regime_20260608T133638/fractonaut_observations.jsonl",
    "/mnt/d/Resonance_Engine/fractonaut_chronicle.jsonl",
]

for p in paths:
    print("=" * 78)
    print(f"FILE: {p}")
    print("=" * 78)
    es = load(p)
    print(f"entries: {len(es)}")
    texts = [text_of(e) for e in es]
    texts = [t for t in texts if t]
    print(f"non-empty: {len(texts)}")
    lens = sorted(len(t) for t in texts)
    if not lens:
        print("(no text)\n"); continue
    print(f"length min/med/max: {lens[0]} / {lens[len(lens)//2]} / {lens[-1]}")

    boiler = sum(1 for t in texts if re.match(r"^[Aa]symmetry", t))
    idle = sum(1 for t in texts if t.lower().startswith("idle"))
    cycle_ref = sum(1 for t in texts if "cycle 11943500" in t)
    market_words = sum(1 for t in texts if re.search(r"\b(market|trade|buy|sell|price|volatility|liquidity|order|volume|spike|absorb|bid|ask|flow)\b", t, re.I))
    pattern_words = sum(1 for t in texts if re.search(r"\b(resembl|similar|recall|earlier|history|like .*(before))\b", t, re.I))
    print(f"start with 'asymmetry':   {boiler}/{len(texts)}")
    print(f"start with 'idle':        {idle}/{len(texts)}")
    print(f"reference cycle 11943500: {cycle_ref}/{len(texts)}")
    print(f"mention market/trade words: {market_words}/{len(texts)}")
    print(f"mention pattern/recall:     {pattern_words}/{len(texts)}")

    pairs = sorted(zip([len(t) for t in texts], texts), reverse=True)
    print("\n--- 5 longest responses ---")
    for L, t in pairs[:5]:
        print(f"[{L} chars] {t[:800]}")
        print()

    novel = [t for t in texts if not re.match(r"^[Aa]symmetry", t) and not t.lower().startswith("idle")]
    print(f"\n--- {len(novel)} non-boilerplate (showing 10) ---")
    for t in novel[:10]:
        print(f"  - {t[:400]}")

    words = Counter()
    for t in texts:
        for w in re.findall(r"[a-zA-Z]+", t.lower()):
            if len(w) >= 5: words[w] += 1
    print("\n--- top 20 words (len>=5) ---")
    for w, c in words.most_common(20):
        print(f"  {c:5d}  {w}")
    print()