import os
import re
import logging
from flask import Flask, request, jsonify, make_response

from presidio_analyzer import AnalyzerEngineProvider

# Configuration du logging
logging.basicConfig(level=logging.INFO,
                    format='%(asctime)s - %(name)s - %(levelname)s - %(message)s')
logger = logging.getLogger(__name__)

app = Flask(__name__)

# Initialisation du moteur Presidio via Provider
analyzer = None
try:
    logger.info("--- Presidio Analyzer Service Starting ---")
    CONFIG_FILE_PATH = os.environ.get("PRESIDIO_ANALYZER_CONFIG_FILE", "conf/default.yaml")
    provider = AnalyzerEngineProvider(analyzer_engine_conf_file=CONFIG_FILE_PATH)
    analyzer = provider.create_engine()
    logger.info(f"Analyzer ready. Supported languages: {analyzer.supported_languages}")
except Exception as e:
    logger.exception("Error during AnalyzerEngine initialization.")
    analyzer = None

# Regex stricte pour détecter l'IBAN (lettres majuscules ou minuscules, chiffres, espaces optionnels)
IBAN_REGEX = re.compile(r"\b[A-Z]{2}[0-9]{2}(?:\s?[A-Z0-9]{4}){4,7}\b", re.IGNORECASE)

# Labels / titres à exclure de l'anonymisation (en minuscules pour comparaison simple)
IGNORE_LABELS = {
    "témoins",
    "témoins clés",
    "coordonnées",
    "coordonnées bancaires",
    "contexte financier",
    "données sensibles",
    "contexte",
    # Ajoute d'autres labels si besoin
}

def normalize_label(txt):
    return txt.strip().lower()

@app.route('/analyze', methods=['POST'])
def analyze_text():
    if not analyzer:
        return jsonify({"error": "Analyzer engine is not available. Check startup logs."}), 500

    try:
        data = request.get_json(force=True)
        text_to_analyze = data.get("text", "")
        language = data.get("language", "fr")

        if not text_to_analyze:
            return jsonify({"error": "text field is missing or empty"}), 400

        results = analyzer.analyze(
            text=text_to_analyze,
            language=language
        )

        filtered_results = []
        for res in results:
            ent_text = text_to_analyze[res.start:res.end].strip()
            ent_text_norm = normalize_label(ent_text)

            # 1. Ignorer les entités correspondant aux labels/titres à préserver
            if ent_text_norm in IGNORE_LABELS:
                logger.debug(f"Skipping anonymization of label: '{ent_text}'")
                continue

            # 2. Si entité de type IBAN, recadrer strictement sur le match IBAN
            if res.entity_type == "IBAN":
                match = IBAN_REGEX.search(ent_text)
                if match:
                    true_iban = match.group(0)
                    start_offset = ent_text.find(true_iban)
                    if start_offset != -1:
                        old_start, old_end = res.start, res.end
                        res.start += start_offset
                        res.end = res.start + len(true_iban)
                        ent_text = true_iban
                        logger.debug(
                            f"Adjusted IBAN span from ({old_start}-{old_end}) to ({res.start}-{res.end}): '{ent_text}'"
                        )
                    else:
                        logger.warning(f"Cannot find IBAN substring inside entity text: '{ent_text}'")
                else:
                    # Si pas de correspondance IBAN, on ignore cette entité (évite faux positifs)
                    logger.warning(f"Entity IBAN does not match IBAN regex: '{ent_text}'")
                    continue  # Ne pas ajouter l'entité

            filtered_results.append(res)

        # Conversion en dict JSON pour renvoyer la réponse
        response_data = [res.to_dict() for res in filtered_results]
        return make_response(jsonify(response_data), 200)

    except Exception as e:
        logger.exception(f"Error during analysis for language '{language}'.")
        if "No matching recognizers" in str(e):
            return jsonify({"error": f"No recognizers available for language '{language}'."}), 400
        return jsonify({"error": str(e)}), 500


if __name__ == '__main__':
    app.run(host='0.0.0.0', port=5001)