#!/usr/bin/env python3
"""Compare one extension manifest object parts with an XML export tree.

The script is intentionally evidence-first. It reports mechanical facts:
manifest object_id parts, ConfigCAS keys, decoded payload structure, embedded
base64 blobs, and exact/contains matches against files from an XML export.
It does not assign semantic names to suffixes.
"""

from __future__ import annotations

import argparse
import base64
import hashlib
import json
import re
from pathlib import Path
from typing import Any

from inspect_1c_sql_files import GUID_RE, Lexer, Parser, collect_strings, tree_shape, try_decode, try_decompress


BASE64_RE = re.compile(r"[A-Za-z0-9+/]{40,}={0,2}")
BSL_MARKERS = ("&На", "Процедура ", "Функция ", "#Область", "#КонецОбласти")
HTML_MARKERS = ("<!DOCTYPE", "<html", "<HTML", "<body", "<BODY")
STREAM_HEADER_RE = re.compile(rb"\r\n([0-9a-f]{8}) ([0-9a-f]{8}) 7fffffff \r\n")


def scalar(node: Any) -> str:
    if isinstance(node, dict) and node.get("type") in {"atom", "string"}:
        return str(node.get("value") or "")
    return ""


def suffix_of(object_id: str) -> str:
    parts = object_id.split(".", 1)
    return "" if len(parts) == 1 else "." + parts[1]


def sha1_hex(data: bytes) -> str:
    return hashlib.sha1(data).hexdigest()


def decode_text(data: bytes) -> tuple[str | None, str | None]:
    if data.startswith(b"\xef\xbb\xbf"):
        try:
            return data.decode("utf-8-sig"), "utf-8-sig"
        except UnicodeDecodeError:
            pass
    return try_decode(data)


def decode_payload_text(data: bytes) -> tuple[str | None, str | None, int]:
    if data.startswith(b"\xef\xbb\xbf"):
        text, encoding = decode_text(data)
        return text, encoding, 0
    marker = data.find(b"\xef\xbb\xbf")
    if marker >= 0:
        try:
            return data[marker:].decode("utf-8-sig"), "utf-8-sig", marker
        except UnicodeDecodeError:
            pass
    text, encoding = try_decode(data)
    return text, encoding, 0


def payload_markers(data: bytes) -> list[str]:
    markers = []
    if data.startswith(b"MOXCEL"):
        markers.append("MOXCEL")
    if data.startswith(b"\xef\xbb\xbf") or b"\xef\xbb\xbf" in data[:256]:
        markers.append("utf8_bom")
    if STREAM_HEADER_RE.search(data):
        markers.append("stream_headers")
    return markers


def load_xml_files(paths: list[Path]) -> list[dict[str, Any]]:
    files: list[dict[str, Any]] = []
    seen: set[Path] = set()
    for root in paths:
        if not root.exists():
            continue
        candidates = [root] if root.is_file() else [item for item in root.rglob("*") if item.is_file()]
        for path in candidates:
            resolved = path.resolve()
            if resolved in seen:
                continue
            seen.add(resolved)
            data = path.read_bytes()
            text, encoding = decode_text(data)
            files.append(
                {
                    "path": str(path),
                    "name": path.name,
                    "relative_hint": str(path),
                    "bytes": len(data),
                    "sha1": sha1_hex(data),
                    "text": text,
                    "encoding": encoding,
                }
            )
    return files


def collect_atoms(value: Any) -> list[str]:
    atoms: list[str] = []

    def walk(node: Any) -> None:
        if isinstance(node, dict) and node.get("type") == "atom":
            atoms.append(str(node.get("value") or ""))
        if isinstance(node, dict):
            for child in node.get("items") or []:
                walk(child)

    walk(value)
    return atoms


def collect_base64_blocks(value: Any) -> list[str]:
    blocks: list[str] = []

    def walk(node: Any) -> None:
        if isinstance(node, dict) and node.get("type") == "list":
            items = node.get("items") or []
            if items and scalar(items[0]) == "#base64":
                chunks = [scalar(item) for item in items[1:] if BASE64_RE.fullmatch(scalar(item))]
                if chunks:
                    blocks.append("".join(chunks))
            for child in items:
                walk(child)

    walk(value)
    return blocks


def decode_base64_atoms(atoms: list[str]) -> list[dict[str, Any]]:
    decoded: list[dict[str, Any]] = []
    for value in atoms:
        if not BASE64_RE.fullmatch(value):
            continue
        try:
            data = base64.b64decode(value, validate=True)
        except Exception:
            continue
        if not data:
            continue
        text, encoding = decode_text(data)
        text_preview = ""
        if text:
            text_preview = text.replace("\x00", "")[:300]
        decoded.append(
            {
                "atom_length": len(value),
                "bytes": len(data),
                "sha1": sha1_hex(data),
                "encoding": encoding,
                "text_preview": text_preview,
                "has_bsl_marker": bool(text and any(marker in text for marker in BSL_MARKERS)),
                "has_html_marker": bool(text and any(marker in text for marker in HTML_MARKERS)),
                "bytes_base64": value[:120],
                "data": data,
                "text": text,
            }
        )
    return decoded


def decode_base64_blocks(blocks: list[str]) -> list[dict[str, Any]]:
    decoded: list[dict[str, Any]] = []
    for value in blocks:
        try:
            data = base64.b64decode(value, validate=True)
        except Exception:
            continue
        text, encoding = decode_text(data)
        decoded.append(
            {
                "block_length": len(value),
                "bytes": len(data),
                "sha1": sha1_hex(data),
                "encoding": encoding,
                "text_preview": (text or "").replace("\x00", "")[:500],
                "has_bsl_marker": bool(text and any(marker in text for marker in BSL_MARKERS)),
                "has_html_marker": bool(text and any(marker in text for marker in HTML_MARKERS)),
                "data": data,
                "text": text,
            }
        )
    return decoded


def match_blob(blob: bytes, text: str | None, xml_files: list[dict[str, Any]]) -> dict[str, Any]:
    exact = []
    normalized_equal = []
    contains = []
    blob_sha1 = sha1_hex(blob)
    normalized_text = text.replace("\r\n", "\n").strip() if text else None
    for item in xml_files:
        file_text = item.get("text")
        data = Path(item["path"]).read_bytes()
        if item["sha1"] == blob_sha1:
            exact.append({"path": item["path"], "match": "sha1"})
        elif normalized_text and file_text and normalized_text == file_text.replace("\r\n", "\n").strip():
            normalized_equal.append({"path": item["path"], "match": "normalized_text_equal"})
        elif len(blob) >= 24 and blob in data:
            contains.append({"path": item["path"], "match": "bytes_contains"})
        elif text and file_text and len(text.strip()) >= 24 and text.strip() in file_text:
            contains.append({"path": item["path"], "match": "text_contains"})
        elif text and file_text and len(file_text.strip()) >= 24 and file_text.strip() in text:
            contains.append({"path": item["path"], "match": "payload_contains_file_text"})
    return {"exact": exact, "normalized_equal": normalized_equal, "contains": contains}


def extract_stream_blocks(payload: bytes) -> list[dict[str, Any]]:
    blocks: list[dict[str, Any]] = []
    for match in STREAM_HEADER_RE.finditer(payload):
        declared_1 = int(match.group(1), 16)
        declared_2 = int(match.group(2), 16)
        start = match.end()
        size = declared_2
        if size <= 0 or start + size > len(payload):
            continue
        data = payload[start : start + size]
        text, encoding = decode_text(data)
        blocks.append(
            {
                "header_offset": match.start(),
                "data_offset": start,
                "declared_1": declared_1,
                "declared_2": declared_2,
                "bytes": len(data),
                "sha1": sha1_hex(data),
                "encoding": encoding,
                "text_preview": (text or "").replace("\x00", "")[:500],
                "has_bsl_marker": bool(text and any(marker in text for marker in BSL_MARKERS)),
                "has_html_marker": bool(text and any(marker in text for marker in HTML_MARKERS)),
                "data": data,
                "text": text,
            }
        )
    return blocks


def match_strings(strings: list[str], xml_files: list[dict[str, Any]], *, limit: int = 80) -> list[dict[str, Any]]:
    hits: list[dict[str, Any]] = []
    for string in strings:
        if len(string.strip()) < 4:
            continue
        paths = []
        for item in xml_files:
            text = item.get("text") or ""
            if string in text:
                paths.append(item["path"])
                if len(paths) >= 8:
                    break
        if paths:
            hits.append({"string": string[:200], "paths": paths})
        if len(hits) >= limit:
            break
    return hits


def match_guids(text: str, xml_files: list[dict[str, Any]], *, limit: int = 200) -> list[dict[str, Any]]:
    hits: list[dict[str, Any]] = []
    for guid in sorted(set(match.lower() for match in GUID_RE.findall(text))):
        paths = []
        for item in xml_files:
            file_text = (item.get("text") or "").lower()
            if guid in file_text:
                paths.append(item["path"])
                if len(paths) >= 8:
                    break
        if paths:
            hits.append({"guid": guid, "paths": paths})
        if len(hits) >= limit:
            break
    return hits


def parse_cas_payload(path: Path) -> dict[str, Any]:
    raw = path.read_bytes()
    payload, compression = try_decompress(raw)
    markers = payload_markers(payload)
    if "stream_headers" in markers and not payload.startswith(b"\xef\xbb\xbf"):
        text, encoding, text_offset = None, None, 0
    else:
        text, encoding, text_offset = decode_payload_text(payload)
    report: dict[str, Any] = {
        "cas_file": path.name,
        "bytes": len(raw),
        "payload_bytes": len(payload),
        "payload_sha1": sha1_hex(payload),
        "compression": compression,
        "encoding": encoding,
        "text_offset": text_offset,
        "payload_markers": markers,
        "parse_status": "not_text",
        "text_preview": "",
        "strings": [],
        "base64_blobs": [],
        "base64_blocks": [],
        "stream_blocks": [],
    }
    report["_raw_payload"] = payload
    report["_stream_blocks"] = extract_stream_blocks(payload)
    if text is None:
        return report
    clean = text.replace("\x00", "").replace("\ufeff", "").lstrip("ï»¿п»ї")
    report["text_preview"] = clean[:500]
    report["has_bsl_marker"] = any(marker in clean for marker in BSL_MARKERS)
    report["has_html_marker"] = any(marker in clean for marker in HTML_MARKERS)
    if "{" not in clean:
        report["parse_status"] = "text_no_braces"
        return report
    try:
        parsed = Parser(Lexer(clean[:2_000_000]).tokens()).parse()
    except Exception as exc:
        report["parse_status"] = "parse_error"
        report["parse_error"] = str(exc)
        return report
    report["parse_status"] = "parsed"
    report["shape"] = tree_shape(parsed, max_depth=4)
    strings = collect_strings(parsed, limit=200)
    atoms = collect_atoms(parsed)
    blocks = collect_base64_blocks(parsed)
    report["strings"] = strings
    if isinstance(parsed, dict):
        items = parsed.get("items") or []
        report["root_type"] = parsed.get("type")
        report["root_len"] = len(items)
        report["root_marker"] = scalar(items[0]) if items else ""
    report["_raw_payload"] = payload
    report["_clean_text"] = clean
    report["_base64_decoded"] = decode_base64_atoms(atoms)
    report["_base64_blocks_decoded"] = decode_base64_blocks(blocks)
    return report


def load_manifest_entries(manifest_path: Path, object_guid: str) -> list[dict[str, Any]]:
    manifest = json.loads(manifest_path.read_text(encoding="utf-8"))
    target = object_guid.lower()
    entries = []
    for entry in manifest.get("entries") or []:
        object_id = str(entry.get("object_id") or "").lower()
        if object_id == target or object_id.startswith(target + "."):
            entries.append(entry)
    entries.sort(key=lambda item: (suffix_of(str(item["object_id"])), str(item["object_id"])))
    return entries


def public_payload_report(payload_report: dict[str, Any], xml_files: list[dict[str, Any]]) -> dict[str, Any]:
    raw_payload = payload_report.pop("_raw_payload", b"")
    clean_text = payload_report.pop("_clean_text", "")
    decoded = payload_report.pop("_base64_decoded", [])
    decoded_blocks = payload_report.pop("_base64_blocks_decoded", [])
    stream_blocks = payload_report.pop("_stream_blocks", [])
    payload_report["payload_matches"] = match_blob(raw_payload, clean_text, xml_files)
    payload_report["string_matches"] = match_strings(payload_report.get("strings") or [], xml_files)
    payload_report["guid_matches"] = match_guids(clean_text, xml_files)
    public_decoded = []
    for blob in decoded:
        data = blob.pop("data")
        text = blob.pop("text")
        blob["matches"] = match_blob(data, text, xml_files)
        public_decoded.append(blob)
    payload_report["base64_blobs"] = public_decoded
    public_blocks = []
    for blob in decoded_blocks:
        data = blob.pop("data")
        text = blob.pop("text")
        blob["matches"] = match_blob(data, text, xml_files)
        public_blocks.append(blob)
    payload_report["base64_blocks"] = public_blocks
    public_streams = []
    for block in stream_blocks:
        data = block.pop("data")
        text = block.pop("text")
        block["matches"] = match_blob(data, text, xml_files)
        public_streams.append(block)
    payload_report["stream_blocks"] = public_streams
    payload_report["strings"] = (payload_report.get("strings") or [])[:80]
    return payload_report


def main() -> int:
    parser = argparse.ArgumentParser(description="Compare one manifest object's CAS parts with XML files.")
    parser.add_argument("--manifest", type=Path, required=True)
    parser.add_argument("--cas-dir", type=Path, required=True)
    parser.add_argument("--object-guid", required=True)
    parser.add_argument("--xml-path", type=Path, action="append", default=[])
    parser.add_argument("--output", type=Path, required=True)
    args = parser.parse_args()

    xml_files = load_xml_files(args.xml_path)
    entries = load_manifest_entries(args.manifest, args.object_guid)
    parts = []
    for entry in entries:
        cas_key = entry["cas_key"]
        cas_path = Path(entry.get("cas_path") or args.cas_dir / cas_key)
        if not cas_path.is_file():
            cas_path = args.cas_dir / cas_key
        payload = parse_cas_payload(cas_path) if cas_path.is_file() else {"parse_status": "missing_cas"}
        parts.append(
            {
                "object_id": entry["object_id"],
                "suffix": suffix_of(entry["object_id"]),
                "cas_key": cas_key,
                "cas_path": str(cas_path),
                "payload": public_payload_report(payload, xml_files) if cas_path.is_file() else payload,
            }
        )

    report = {
        "schema": "onec_manifest_object_part_compare.v1",
        "object_guid": args.object_guid.lower(),
        "manifest": str(args.manifest),
        "cas_dir": str(args.cas_dir),
        "xml_paths": [str(path) for path in args.xml_path],
        "xml_file_count": len(xml_files),
        "xml_files": [
            {
                "path": item["path"],
                "bytes": item["bytes"],
                "sha1": item["sha1"],
                "encoding": item["encoding"],
            }
            for item in xml_files
        ],
        "part_count": len(parts),
        "parts": parts,
    }
    args.output.parent.mkdir(parents=True, exist_ok=True)
    args.output.write_text(json.dumps(report, ensure_ascii=False, indent=2) + "\n", encoding="utf-8")
    print(json.dumps({"output": str(args.output), "parts": len(parts), "xml_files": len(xml_files)}, ensure_ascii=False))
    return 0


if __name__ == "__main__":
    raise SystemExit(main())