"""Configuration loaded from Infisical or central .env file.

Priority: Infisical → environment variables → .env file
"""

import os
from pathlib import Path

from dotenv import load_dotenv

# Load from central .env or override path
dotenv_path = os.environ.get("DOTENV_PATH", str(Path.home() / ".env"))
load_dotenv(dotenv_path)

# Try loading from Infisical if configured
INFISICAL_TOKEN = os.environ.get("INFISICAL_TOKEN", "")
if INFISICAL_TOKEN:
    try:
        from infisical_sdk import InfisicalSDKClient
        _client = InfisicalSDKClient(token=INFISICAL_TOKEN)
        _secrets = _client.get_all_secrets(
            environment=os.environ.get("INFISICAL_ENV", "production"),
            project_id=os.environ.get("INFISICAL_PROJECT_ID", ""),
        )
        for s in _secrets:
            os.environ.setdefault(s.secret_key, s.secret_value)
    except ImportError:
        pass  # Infisical SDK not installed — use .env
    except Exception:
        pass  # Infisical unreachable — fall back to .env

# PostgreSQL
POSTGRES_URL = os.environ.get(
    "POSTGRES_URL",
    f"postgres://{os.environ.get('POSTGRES_USER', 'legal_ai')}:"
    f"{os.environ.get('POSTGRES_PASSWORD', '')}@"
    f"{os.environ.get('POSTGRES_HOST', '127.0.0.1')}:"
    f"{os.environ.get('POSTGRES_PORT', '5433')}/"
    f"{os.environ.get('POSTGRES_DB', 'legal_ai')}",
)

# Redis
REDIS_URL = os.environ.get("REDIS_URL", "redis://127.0.0.1:6380/0")

# Voyage AI
VOYAGE_API_KEY = os.environ.get("VOYAGE_API_KEY", "")
VOYAGE_MODEL = os.environ.get("VOYAGE_MODEL", "voyage-law-2")
VOYAGE_DIMENSIONS = 1024

# Rerank — cross-encoder second-stage. Off by default; flip with env to
# enable across all semantic search tools (search_decisions,
# search_case_documents, find_similar_cases, search_precedent_library).
VOYAGE_RERANK_MODEL = os.environ.get("VOYAGE_RERANK_MODEL", "rerank-2")
VOYAGE_RERANK_ENABLED = (
    os.environ.get("VOYAGE_RERANK_ENABLED", "false").lower() == "true"
)
# How many candidates to fetch from bi-encoder before reranking.
# 50 was the depth used in the POC; balances recall vs rerank cost.
VOYAGE_RERANK_FETCH_K = int(os.environ.get("VOYAGE_RERANK_FETCH_K", "50"))

# Multimodal — page-image embeddings via voyage-multimodal-3. Off by
# default; flip with env to enable per-page image embedding during
# ingestion + hybrid (text+image) ranking at search time. POC #3
# validated on a 89-page appraisal PDF (38s, 312K tokens, recovered
# table structure + image-only scanned pages that text-OCR misses).
MULTIMODAL_ENABLED = (
    os.environ.get("MULTIMODAL_ENABLED", "false").lower() == "true"
)
MULTIMODAL_MODEL = os.environ.get("MULTIMODAL_MODEL", "voyage-multimodal-3")
# Render DPI for the image fed to the embedder. POC used 144 — sweet
# spot between embedding quality and tokens/page (144 ≈ 3.5K tok/page).
MULTIMODAL_DPI = int(os.environ.get("MULTIMODAL_DPI", "144"))
# Separate, lower DPI for the JPEG thumbnail saved to disk for UI
# preview. ~96dpi → ~20KB/page; ingestion-time, no re-render at view.
MULTIMODAL_THUMB_DPI = int(os.environ.get("MULTIMODAL_THUMB_DPI", "96"))
# Hybrid merge: Reciprocal Rank Fusion (RRF) bias for the *text* side.
# voyage-3 cosine scores (~0.4-0.5) and voyage-multimodal-3 scores
# (~0.20-0.25) live on different scales; a direct weighted sum lets
# text always dominate. RRF is rank-based and robust to that. The
# weight here biases the contribution of each side: 0.5 = balanced
# (vanilla RRF), >0.5 favours text, <0.5 favours image. Tunable per
# env without redeploy.
MULTIMODAL_TEXT_WEIGHT = float(
    os.environ.get("MULTIMODAL_TEXT_WEIGHT", "0.5")
)
# RRF damping constant. Standard literature value is 60: lower values
# concentrate weight at top ranks; higher values flatten the curve.
MULTIMODAL_RRF_K = int(os.environ.get("MULTIMODAL_RRF_K", "60"))

# Halacha extraction — auto-approve threshold. Halachot with extractor
# confidence >= this value are inserted with review_status='approved'
# instead of 'pending_review' (so they immediately appear in
# search_precedent_library). Set to a value > 1.0 to disable auto-approval.
# 0.80 baseline: 89% of historical extractions land here, manual spot-check
# of 10 random samples confirmed quality. Tunable via env if drift is
# observed (e.g. raise to 0.90 if false-positives appear).
HALACHA_AUTO_APPROVE_THRESHOLD = float(
    os.environ.get("HALACHA_AUTO_APPROVE_THRESHOLD", "0.80")
)

# Google Cloud Vision (OCR for scanned PDFs)
GOOGLE_CLOUD_VISION_API_KEY = os.environ.get("GOOGLE_CLOUD_VISION_API_KEY", "")

# Data directory
DATA_DIR = Path(os.environ.get("DATA_DIR", str(Path.home() / "legal-ai" / "data")))
TRAINING_DIR = DATA_DIR / "training"
EXPORTS_DIR = DATA_DIR / "exports"  # legacy exports only

# Cases directory — flat structure: data/cases/{case_number}/
CASES_DIR = DATA_DIR / "cases"


def find_case_dir(case_number: str) -> Path:
    """Return the case directory for a given case number."""
    return CASES_DIR / case_number

# Chunking parameters
CHUNK_SIZE_TOKENS = 600
CHUNK_OVERLAP_TOKENS = 100

# External service allowlist — case materials may ONLY be sent to these domains
ALLOWED_EXTERNAL_SERVICES = {
    "api.voyageai.com",         # Voyage AI (embeddings)
    "vision.googleapis.com",    # Google Cloud Vision (OCR)
}

# Audit
AUDIT_ENABLED = os.environ.get("AUDIT_ENABLED", "true").lower() == "true"


# ── Utility ───────────────────────────────────────────────────────

def parse_llm_json(raw: str):
    """Parse JSON from LLM response, handling markdown wrapping and truncation.

    Handles:
    1. Markdown ```json ... ``` code blocks
    2. Extra text before/after JSON
    3. Truncated JSON (missing closing brackets) — attempts recovery
    """
    import json
    import re
    raw = raw.strip()
    # Strip markdown code blocks
    raw = re.sub(r"^```(?:json)?\s*\n?", "", raw)
    raw = re.sub(r"\n?\s*```\s*$", "", raw)
    # Try direct parse first
    try:
        return json.loads(raw)
    except json.JSONDecodeError:
        pass
    # Try to find JSON object or array
    for pattern in [r"\{.*\}", r"\[.*\]"]:
        match = re.search(pattern, raw, re.DOTALL)
        if match:
            try:
                return json.loads(match.group())
            except json.JSONDecodeError:
                continue
    # Attempt truncated JSON recovery:
    # Find the start of JSON, then try closing open brackets
    for opener, closer in [("[", "]"), ("{", "}")]:
        start = raw.find(opener)
        if start < 0:
            continue
        fragment = raw[start:]
        # Try progressively removing trailing partial content and closing
        # Look for the last complete item (ending with }, or ])
        for end_pattern in [r'.*\}(?=\s*,?\s*$)', r'.*\](?=\s*,?\s*$)', r'.*"(?=\s*$)']:
            pass  # fallback below
        # Simple approach: find last complete JSON item boundary
        # For arrays: find last "}" and close the array
        if opener == "[":
            last_brace = fragment.rfind("}")
            if last_brace > 0:
                truncated = fragment[:last_brace + 1] + "]"
                try:
                    return json.loads(truncated)
                except json.JSONDecodeError:
                    pass
        # For objects: find last complete key-value
        if opener == "{":
            last_brace = fragment.rfind("}")
            if last_brace > 0:
                # Check if this closes a nested object — try adding outer close
                truncated = fragment[:last_brace + 1]
                # Count unclosed braces
                open_count = truncated.count("{") - truncated.count("}")
                truncated += "}" * open_count
                try:
                    return json.loads(truncated)
                except json.JSONDecodeError:
                    pass
    return None