"""Local privacy gateway. Personal generative egress is not allowed to skip this layer."""
from __future__ import annotations

from dataclasses import dataclass, field
from datetime import datetime, timezone
from typing import Any

import re

from entity_detect import detect_and_remember
from identity_store import KINSHIP, is_maskable_label, list_mappings
from pronoun_bind import bind_user_lines
from providers import ProviderError, complete_chat, generate_provider

MAX_EGRESS_CHARS = 24000
debug_last: dict[str, Any] | None = None
debug_calls = 0


class PrivacyGatewayError(Exception):
    def __init__(self, code: str, message: str, status_code: int = 503):
        super().__init__(message)
        self.code = code
        self.message = message
        self.status_code = status_code


@dataclass
class GatewayRequest:
    prompt_id: str | None
    purpose: str
    data_class: str
    payload: dict[str, Any] = field(default_factory=dict)
    profile_id: str | None = None


@dataclass
class GatewayResult:
    allowed: bool
    reason: str
    provider: str | None = None
    content: str | None = None
    checked_at: str = field(default_factory=lambda: datetime.now(timezone.utc).isoformat())


ALLOWED_CLASSES = {"A", "B", "C"}
LOCAL_ONLY_CLASS = "A"


def reset_debug() -> None:
    global debug_last, debug_calls
    debug_last = None
    debug_calls = 0


def inspect(request: GatewayRequest) -> GatewayResult:
    """Fail closed unless a policy-complete generative provider is configured."""
    data_class = (request.data_class or "").upper()
    if data_class not in ALLOWED_CLASSES:
        return GatewayResult(allowed=False, reason="unknown_data_class")
    if data_class == LOCAL_ONLY_CLASS:
        return GatewayResult(allowed=False, reason="class_a_never_leaves_local_zone")
    config = generate_provider()
    if not config:
        return GatewayResult(allowed=False, reason="no_egress_provider_configured")
    return GatewayResult(allowed=True, reason="policy_ok", provider=config.name)


def _minimize(text: str) -> str:
    cleaned = (text or "").strip()
    if len(cleaned) <= MAX_EGRESS_CHARS:
        return cleaned
    return cleaned[:MAX_EGRESS_CHARS]


_LETTER = r"A-Za-zÄÖÜäöüß"
_WORD = re.compile(rf"[{_LETTER}]+")
THING_GOVERNORS = {
    "esse", "essen", "isst", "aß", "aßest", "gegessen",
    "koche", "kochen", "kochte", "gekocht",
    "trinke", "trinken", "trank", "getrunken",
    "kaufe", "kaufen", "kaufte", "gekauft",
    "bestelle", "bestellen", "bestellte",
    "hole", "holen", "holte",
}


def _words_before(text: str, index: int, n: int = 4) -> list[str]:
    return [word.lower() for word in _WORD.findall(text[:index])[-n:]]


def _words_after(text: str, index: int, n: int = 3) -> list[str]:
    return [word.lower() for word in _WORD.findall(text[index:])[:n]]


def _is_person_token(token: str) -> bool:
    raw = (token or "").strip().upper()
    raw = raw[2:-2] if raw.startswith("[[") and raw.endswith("]]") else raw
    return raw.startswith("PERSON:")


def _mask_person_hit(text: str, start: int, end: int) -> bool:
    prev = _words_before(text, start)
    nxt = _words_after(text, end)
    if prev and prev[-1] in KINSHIP:
        return True
    if any(word in THING_GOVERNORS for word in prev + nxt):
        return False
    return True


def _mask_body(text: str, mappings: list[dict]) -> str:
    masked = text
    for item in sorted(mappings, key=lambda row: len(row.get("local_label") or ""), reverse=True):
        label = (item.get("local_label") or "").strip()
        token = (item.get("token") or "").strip()
        if not label or not token or not is_maskable_label(label):
            continue
        placeholder = token if token.startswith("[[") else f"[[{token}]]"
        pattern = re.compile(
            rf"(?<![{_LETTER}]){re.escape(label)}(?![{_LETTER}])",
            re.IGNORECASE,
        )
        person = _is_person_token(token)

        def repl(match: re.Match, *, _person=person, _ph=placeholder) -> str:
            if _person and not _mask_person_hit(masked, match.start(), match.end()):
                return match.group(0)
            return _ph

        masked = pattern.sub(repl, masked)
    return masked


def _mask(text: str, mappings: list[dict], *, personal_lines_only: bool = False) -> str:
    if not personal_lines_only:
        return _mask_body(text, mappings)
    parts: list[str] = []
    for line in (text or "").splitlines(keepends=True):
        raw = line[:-1] if line.endswith("\n") else line
        newline = "\n" if line.endswith("\n") else ""
        if raw.startswith("user:") or raw.startswith("assistant:"):
            parts.append(_mask_body(raw, mappings) + newline)
        else:
            parts.append(line)
    return "".join(parts)


def mask_for_egress(rendered: str, mappings: list[dict], purpose: str) -> str:
    """Mask names. Bind pronouns to person tokens only in the dialogue turn.

    Journal generation keeps sie/er/ihr so demask does not turn every reference into the name.
    """
    masked = _mask(rendered, mappings, personal_lines_only=purpose == "dialogue_turn")
    if purpose == "dialogue_turn":
        return bind_user_lines(masked)
    return masked


def _demask(text: str, mappings: list[dict]) -> str:
    result = text or ""
    for item in mappings:
        label = (item.get("local_label") or "").strip()
        token = (item.get("token") or "").strip()
        if not label or not token:
            continue
        placeholder = token if token.startswith("[[") else f"[[{token}]]"
        result = result.replace(placeholder, label)
    return result


def _validate_response(content: str, mappings: list[dict] | None = None) -> str:
    text = (content or "").strip()
    if not text:
        raise PrivacyGatewayError("empty_provider_response", "Der Provider lieferte keine Antwort.")
    leaked: list[str] = []
    for item in mappings or []:
        label = (item.get("local_label") or "").strip()
        if not label or not is_maskable_label(label):
            continue
        if re.search(
            rf"(?<![{_LETTER}]){re.escape(label)}(?![{_LETTER}])",
            text,
            re.IGNORECASE,
        ):
            leaked.append(label)
    if leaked:
        raise PrivacyGatewayError(
            "response_validation_failed",
            "Antwort enthielt Klartext-Identität vor der Demaskierung.",
        )
    return text


def _fake_complete(purpose: str, rendered: str) -> str:
    if purpose == "profile_review":
        return (
            '{"kind":"kansho.profile_analysis_result","format_version":1,"target":"writing",'
            '"evidence_basis":["kansho_sources"],'
            '"changes":[{"layer":"trait","key":"dry_humor","slug":"dry_humor","facet_key":"autobiographical_journal",'
            '"action":"add","label":"Trockener Humor",'
            '"proposed_value":"gelegentlich trocken, nie aufgesetzt",'
            '"evidence_basis":["kansho_sources"],'
            '"rationale":"Fake-Review: Journal-Evidenz trägt Humor in autobiographical_journal, nicht als globalen Core.",'
            '"evidence_ids":[],"exemplars":[{"excerpt":"haha das war irgendwie lustig","role":"exemplar","evidence_basis":["kansho_sources"]}]}]}'
        )
    if purpose == "journal_generate":
        lines = []
        for line in rendered.splitlines():
            if line.startswith("user:"):
                lines.append(line[5:].strip())
        body = " ".join(part for part in lines if part) or "Ein stiller Tag."
        return f"Ein Tag\n\n{body}"
    if "Keine Interviewfrage" in rendered:
        return '{"operation":"fortfuehren","impulse":"Der letzte Faden bleibt offen."}'
    impulse = (
        "Was davon möchtest du vertiefen, ohne Unerwähntes als nicht geschehen zu behandeln?"
        if ("nicht geschehen" in rendered or "nicht erwähnt" in rendered)
        else "Was davon möchtest du festhalten?"
    )
    return '{"operation":"erleben_vertiefen","impulse":"' + impulse + '"}'


def complete_model(messages: list[dict], policy: dict[str, Any]) -> str:
    config = generate_provider()
    if not config:
        raise PrivacyGatewayError("no_egress_provider_configured", "Es ist kein Egress-Provider konfiguriert.")
    if config.mode == "fake":
        rendered = "\n".join(item.get("content") or "" for item in messages)
        return _fake_complete(policy.get("purpose") or "", rendered)
    try:
        return complete_chat(config, messages, timeout=60)
    except ProviderError as exc:
        raise PrivacyGatewayError(exc.code, exc.message, exc.status_code) from exc


def complete(request: GatewayRequest) -> GatewayResult:
    global debug_last, debug_calls
    result = inspect(request)
    if not result.allowed:
        raise PrivacyGatewayError(
            result.reason,
            "Persönlicher KI-Aufruf wurde vom Privacy Gateway blockiert. "
            "Es ist kein Egress-Provider konfiguriert."
            if result.reason == "no_egress_provider_configured"
            else "Persönlicher KI-Aufruf wurde vom Privacy Gateway blockiert.",
        )
    rendered = _minimize(str((request.payload or {}).get("rendered") or ""))
    source_text = str((request.payload or {}).get("source_text") or rendered)
    detect_name = None
    try:
        mappings, detect_name, detect_note = detect_and_remember(request.profile_id, source_text)
    except ProviderError as exc:
        raise PrivacyGatewayError(exc.code, exc.message, exc.status_code) from exc
    if not mappings and request.profile_id:
        mappings = list_mappings(request.profile_id)
    masked = mask_for_egress(rendered, mappings, request.purpose)
    debug_calls += 1
    config = generate_provider()
    debug_last = {
        "purpose": request.purpose,
        "layer": {
            "journal_generate": "journalentwurf",
            "profile_review": "profilreview",
        }.get(request.purpose, "dialogzug"),
        "data_class": request.data_class,
        "prompt_slug": (request.payload or {}).get("prompt_slug"),
        "rendered": rendered,
        "masked": masked,
        "mask_input": source_text,
        "provider": result.provider,
        "model": config.model if config else None,
        "detect_provider": detect_name,
        "detect_note": detect_note,
        "mapping_count": len(mappings),
    }
    raw = complete_model(
        [{"role": "user", "content": masked}],
        {"purpose": request.purpose, "zdr": True, "no_train": True},
    )
    validated = _validate_response(raw, mappings)
    result.content = _demask(validated, mappings)
    result.allowed = True
    debug_last["raw"] = validated
    debug_last["reply"] = result.content
    return result


def last_trace() -> dict | None:
    return debug_last


def public_trace(trace: dict | None) -> dict | None:
    """Admin test view. No mapping table, no secrets."""
    if not trace:
        return None
    return {
        "purpose": trace.get("purpose"),
        "layer": trace.get("layer"),
        "data_class": trace.get("data_class"),
        "prompt_slug": trace.get("prompt_slug"),
        "provider": trace.get("provider"),
        "model": trace.get("model"),
        "detect_provider": trace.get("detect_provider"),
        "detect_note": trace.get("detect_note"),
        "mapping_count": trace.get("mapping_count"),
        "intern": trace.get("rendered"),
        "egress": trace.get("masked"),
        "mask_input": trace.get("mask_input"),
        "raw": trace.get("raw"),
        "reply": trace.get("reply"),
    }