#!/usr/bin/env python3
"""certiv-kernel — the file that DEFINES receipt validity, small enough to read.

Verifies a receipt.certiv with NOTHING but the Python standard library:
  python3 certiv-kernel.py <dir>            structural + hash verification (safe)
  python3 certiv-kernel.py <dir> --execute  ALSO RUNS the receipt's commands

  !!  SECURITY  !!  --execute runs author-declared shell commands on THIS
  machine with no isolation. NEVER run --execute on a receipt you do not
  trust. Use a disposable VM, or let a certiv verification node run it in its
  sandbox (network-isolated, non-root, read-only inputs). Without --execute
  the kernel only reads and hashes files under <dir>; it still refuses any
  path that escapes <dir> (absolute, .., symlink, or non-regular file).

This is a deliberately independent implementation of the receipt rules
(04_RECEIPTS): it shares no code with the certiv tool, so agreement between
the two is evidence, not tautology. If this file and the big tool ever
disagree, believe this file, then file a bug.

Exit code 0 = every check passed. Anything else = the receipt names what broke.
"""
import hashlib
import json
import os
import subprocess
import sys
from pathlib import Path
from stat import S_ISREG


def safe_path(base, rel, must_be_file=True):
    """Resolve rel under base, refusing escapes and non-regular files.
    Returns the path, or None if unsafe/missing (caller fails closed)."""
    base_real = Path(base).resolve()
    p = Path(rel)
    if p.is_absolute() or ".." in p.parts or any(x == "" for x in p.parts):
        return None
    full = (base_real / p).resolve()
    if full != base_real and base_real not in full.parents:
        return None
    if must_be_file:
        if full.is_symlink() or not full.exists():
            return None
        st = os.stat(full, follow_symlinks=False)
        if not S_ISREG(st.st_mode) or not os.access(full, os.R_OK):
            return None
    return full

SCHEMA = "certiv_receipt_v1"
NODE_ONLY = ("verification_result", "verified_at", "node_id")
ENV = {"PATH": "/usr/local/bin:/usr/bin:/bin", "HOME": "/tmp", "LANG": "C.UTF-8",
       "LC_ALL": "C.UTF-8", "TZ": "UTC", "PYTHONHASHSEED": "0", "SOURCE_DATE_EPOCH": "0"}


def sha_text(text):
    return hashlib.sha256(text.encode("utf-8")).hexdigest()


def sha_file(path):
    h = hashlib.sha256()
    with open(path, "rb") as f:
        for chunk in iter(lambda: f.read(1 << 20), b""):
            h.update(chunk)
    return h.hexdigest()


def sha_body(payload, field):
    body = {k: v for k, v in payload.items() if k != field}
    return sha_text(json.dumps(body, sort_keys=True, separators=(",", ":")))


def merkle(hashes):
    level = sorted(hashes)
    if not level:
        return sha_text("")
    while len(level) > 1:
        nxt = [sha_text(level[i] + level[i + 1]) for i in range(0, len(level) - 1, 2)]
        if len(level) % 2 == 1:
            nxt.append(level[-1])
        level = nxt
    return level[0]


def run(command, cwd):
    proc = subprocess.run(["/bin/sh", "-c", command], cwd=str(cwd),
                          env=dict(ENV), capture_output=True, timeout=600)
    return proc.returncode, proc.stdout


def check_execution(claim, cwd, problems):
    cid = claim.get("claim_id")
    if claim.get("kind") == "certificate":
        cert = claim.get("certificate", {})
        cpath = safe_path(cwd, cert.get("path", ""))
        if cpath is None:
            problems.append("unsafe_certificate:%s" % cid)
            return
        if sha_file(cpath) != cert.get("sha256"):
            problems.append("certificate_hash_drift:%s" % cid)
            return
        code, _ = run(cert.get("checker", "false"), cwd)
        if code != 0:
            problems.append("checker_failed:%s" % cid)
        return
    code, stdout = run(claim.get("command", "false"), cwd)
    if code != 0:
        problems.append("exit_%s:%s" % (code, cid))
        return
    expect = claim.get("expect", {})
    if "value" in expect:
        observed = stdout.decode("utf-8", "replace").strip()
        declared = str(expect["value"]).strip()
        if "tolerance" in expect:
            ok = abs(float(observed) - float(declared)) <= float(expect["tolerance"])
        else:
            ok = observed == declared
        if not ok:
            problems.append("value_mismatch:%s:got_%s" % (cid, observed[:40]))
    elif "stdout_sha256" in expect:
        if hashlib.sha256(stdout).hexdigest() != expect["stdout_sha256"]:
            problems.append("stdout_hash_mismatch:%s" % cid)
    else:
        out = safe_path(cwd, expect.get("output_file", ""))
        if out is None or sha_file(out) != expect.get("output_sha256"):
            problems.append("output_hash_mismatch:%s" % cid)


def verify(directory, execute=False):
    cwd = Path(directory)
    problems = []
    path = cwd / "receipt.certiv"
    if not path.is_file():
        return ["receipt_missing"]
    r = json.loads(path.read_text(encoding="utf-8"))

    if r.get("schema") != SCHEMA:
        problems.append("schema_invalid")
    for field in NODE_ONLY:
        if field in r:
            problems.append("author_receipt_carries_node_field:%s" % field)
    if sha_body(r, "receipt_sha256") != r.get("receipt_sha256"):
        problems.append("receipt_sha256_mismatch")

    claims = r.get("claims", [])
    shas = []
    for c in claims:
        cid = c.get("claim_id", "?")
        declared = c.get("claim_sha256", "")
        shas.append(declared)
        if sha_body(c, "claim_sha256") != declared:
            problems.append("claim_sha256_mismatch:%s" % cid)
        for field in NODE_ONLY:
            if field in c:
                problems.append("claim_carries_node_field:%s" % cid)
        for pin in c.get("inputs", []):
            p = safe_path(cwd, pin["path"])
            if p is None:
                problems.append("unsafe_input:%s:%s" % (cid, pin["path"]))
            elif sha_file(p) != pin["sha256"]:
                problems.append("input_hash_drift:%s:%s" % (cid, pin["path"]))
    if merkle(shas) != r.get("claims_merkle_root"):
        problems.append("claims_merkle_root_mismatch")

    cov = r.get("coverage", {})
    verified = [c for c in claims
                if c.get("weight") == "load_bearing" and c.get("status") == "verified_local"]
    total = [c for c in claims if c.get("weight") == "load_bearing"]
    if cov.get("load_bearing_verified_local") != len(verified) or \
       cov.get("load_bearing_total") != len(total):
        problems.append("coverage_misstated")

    if execute and not problems:
        for c in claims:
            if c.get("status") == "verified_local":
                check_execution(c, cwd, problems)
    return problems


def main():
    args = [a for a in sys.argv[1:] if not a.startswith("--")]
    execute = "--execute" in sys.argv
    directory = args[0] if args else "."
    if execute:
        sys.stderr.write(
            "WARNING: --execute runs the receipt's shell commands on this machine "
            "with no isolation. Only do this for receipts you trust.\n")
    problems = verify(directory, execute=execute)
    mode = "hashes+structure+execution" if execute else "hashes+structure"
    if problems:
        for p in problems:
            print("KERNEL FAIL  %s" % p)
        print("kernel verdict: NOT CONFIRMED (%s)" % mode)
        return 1
    print("kernel verdict: receipt CONFIRMED (%s)" % mode)
    return 0


if __name__ == "__main__":
    sys.exit(main())