ROMFASTSQL/proxmox/lxc171-claude-agent/maria-whatsapp-bridge/rag/config.py

"""Configuratie comuna a puntii WhatsApp+RAG pentru Maria (LXC 171 claude-agent).

Citeste ~/.maria-bridge/env (KEY=value, tolerant la comentarii si ghilimele).
Mirrors deliberat conventiile din discord-bridge/config.py, ca sa fie un singur
model de citit pentru cine intretine ambele punti pe acest container.
"""

from __future__ import annotations

import contextlib
import fcntl
import os
import pathlib

# Directorul de baza poate fi mutat in teste prin MARIA_BRIDGE_DIR.
_DEFAULT_DIR = pathlib.Path.home() / ".maria-bridge"

STATE_DIR: pathlib.Path = pathlib.Path(os.environ.get("MARIA_BRIDGE_DIR") or _DEFAULT_DIR)
DOCS_DIR: pathlib.Path = STATE_DIR / "documents"
# Documente care NU vin din Drive. `rclone sync` sterge din DOCS_DIR tot ce nu
# exista in Drive, deci orice fisier pus acolo de mana dispare la urmatorul tur.
# Ce sta aici e in afara oglinzii si supravietuieste.
DOCS_LOCAL_DIR: pathlib.Path = STATE_DIR / "documents-local"
INDEX_FILE: pathlib.Path = STATE_DIR / "rag_index.json"
LOCK_FILE: pathlib.Path = STATE_DIR / ".rag.lock"
LOG_DIR: pathlib.Path = STATE_DIR / "logs"
ENV_FILE: pathlib.Path = STATE_DIR / "env"
AUTH_DIR: pathlib.Path = STATE_DIR / "whatsapp-auth"

_env: dict[str, str] = {}

# Valori implicite. LLM_URL/OLLAMA_URL presupun tunel/proxy local catre backend-ul
# real (vezi docs/maria-whatsapp-rag-prototype.md) — de completat in env dupa caz.
DEFAULTS: dict[str, str] = {
    "BRIDGE_HOST": "127.0.0.1",
    "BRIDGE_PORT": "8099",
    "LLM_URL": "http://127.0.0.1:8091",
    "OLLAMA_URL": "http://127.0.0.1:11434",
    "EMBED_MODEL": "nomic-embed-text",
    "TOP_K": "3",
    "MAX_TOKENS": "250",
    "POLL_INTERVAL_S": "2",
    "TEST_MODE_SELF_CHAT_ONLY": "true",
    # Grupuri in care Maria are voie sa raspunda (JID-uri separate prin virgula).
    # Restul grupurilor sunt ignorate, si de punte, si de consumer.
    "ALLOWED_GROUP_JIDS": "",
    # Firul de discutie (vezi rag/fir.py): cat tine minte, cat tace dupa ce a
    # preluat un om, si de cati participanti e nevoie ca preluarea sa se aplice
    # automat (in self-chat si in grupul de test totul e `fromMe`).
    "FIR_TTL_MIN": "120",
    "FIR_TACERE_MIN": "60",
    "FIR_PRELUARE_MIN_PARTICIPANTI": "2",
    # OCR pentru capturile de ecran primite pe WhatsApp (vezi rag/ocr.py).
    "OCR_LANGS": "ron+eng",
    "OCR_MAX_CHARS": "1500",
    "OCR_TIMEOUT_S": "60",
    # Ordonarea chunk-urilor si pragul de la care se considera ca raspunsul chiar
    # e in documente (vezi rag/rank.py pentru masuratoarea din spatele valorilor).
    "RANK_STRONG_COSINE": "0.70",
    "RANK_WEAK_COSINE": "0.58",
    "RANK_RECALL_N": "12",
    "RANK_MIN_RARE_RATIO": "0.5",
    # Unde pleaca intrebarile fara raspuns: JID de WhatsApp, ex.
    # "40712345678@s.whatsapp.net" sau "1203...@g.us" pentru un grup.
    # Gol = escaladarile se scriu doar in jurnal (~/.maria-bridge/escalations/).
    "SUPPORT_JID": "",
    # Tinta rclone pentru sincronizarea depozitului de documente, ex:
    # "gdrive,root_folder_id=1C4e75zgH1_7ZK-_oBP5ZZBvUPh3iEo1O:" (dosarul
    # document_store din Drive, vazut pe Windows ca D:\GoogleDrive\romfast\document_store).
    # Gol = sincronizare dezactivata, doar upload manual din dashboard.
    "DRIVE_REMOTE": "",
}


def parse_env(text: str) -> dict[str, str]:
    """Parseaza un fisier de tip KEY=value. Nu arunca niciodata."""
    out: dict[str, str] = {}
    for raw in text.splitlines():
        line = raw.strip()
        if not line or line.startswith("#"):
            continue
        if line.startswith("export "):
            line = line[len("export "):].strip()
        if "=" not in line:
            continue
        key, _, val = line.partition("=")
        key = key.strip()
        if not key:
            continue
        val = val.strip()
        if val[:1] not in ("'", '"'):
            cut = val.find(" #")
            if cut >= 0:
                val = val[:cut].rstrip()
        if len(val) >= 2 and val[0] == val[-1] and val[0] in ("'", '"'):
            val = val[1:-1]
        out[key] = val
    return out


def reload(base_dir: str | os.PathLike | None = None) -> dict[str, str]:
    """Recalculeaza caile si reciteste env-ul. Returneaza dictionarul incarcat."""
    global STATE_DIR, DOCS_DIR, DOCS_LOCAL_DIR, INDEX_FILE, LOCK_FILE, LOG_DIR, ENV_FILE, AUTH_DIR, _env
    if base_dir is None:
        base_dir = os.environ.get("MARIA_BRIDGE_DIR") or _DEFAULT_DIR
    STATE_DIR = pathlib.Path(base_dir)
    DOCS_DIR = STATE_DIR / "documents"
    DOCS_LOCAL_DIR = STATE_DIR / "documents-local"
    INDEX_FILE = STATE_DIR / "rag_index.json"
    LOCK_FILE = STATE_DIR / ".rag.lock"
    LOG_DIR = STATE_DIR / "logs"
    ENV_FILE = STATE_DIR / "env"
    AUTH_DIR = STATE_DIR / "whatsapp-auth"
    try:
        _env = parse_env(ENV_FILE.read_text(encoding="utf-8"))
    except OSError:
        _env = {}
    return _env


def get(key: str, default: str | None = None) -> str | None:
    if key in _env:
        return _env[key]
    if key in DEFAULTS:
        return DEFAULTS[key]
    return default


def get_int(key: str, default: int) -> int:
    try:
        return int(get(key, str(default)) or default)
    except (TypeError, ValueError):
        return default


reload()


class Busy(RuntimeError):
    """Alta reindexare e deja in curs."""


@contextlib.contextmanager
def exclusive():
    """Lacat intre PROCESE pentru reindexare (timer vs dashboard vs rulare manuala).

    Reindexarea dureaza minute (embeddings pe CPU). Fara lacat, `maria-sync.timer`
    poate porni peste o sincronizare manuala si ambele scriu acelasi rag_index.json.
    Nu asteptam: a doua rulare se anuleaza curat, fiindca oricum ar reface acelasi
    lucru imediat dupa.
    """
    STATE_DIR.mkdir(parents=True, exist_ok=True)
    fh = open(LOCK_FILE, "w", encoding="utf-8")
    try:
        try:
            fcntl.flock(fh, fcntl.LOCK_EX | fcntl.LOCK_NB)
        except OSError:
            raise Busy("o reindexare e deja in curs")
        fh.write(str(os.getpid()))
        fh.flush()
        yield
    finally:
        fh.close()