99 lines
2.8 KiB
Python
99 lines
2.8 KiB
Python
|
"""
|
||
|
|
Extraction de salle/zone depuis le texte - GMAO College
|
||
|
|
Parse le corps des emails/messages pour identifier les salles mentionnees.
|
||
|
|
"""
|
||
|
|
import re
|
||
|
|
from ..extensions import db
|
||
|
|
from ..core.models.college import Room, Building
|
||
|
|
|
||
|
|
|
||
|
|
def extract_rooms_from_text(text):
|
||
|
|
"""Cherche des noms de salles dans le texte et retourne les Rooms correspondantes.
|
||
|
|
|
||
|
|
Strategie :
|
||
|
|
1. Chercher des patterns comme "salle B01", "B01", "couloir", "cour", etc.
|
||
|
|
2. Fuzzy match avec les noms de salles en DB
|
||
|
|
3. Retourne une liste de Room objects
|
||
|
|
"""
|
||
|
|
if not text:
|
||
|
|
return []
|
||
|
|
|
||
|
|
text_lower = text.lower()
|
||
|
|
rooms = Room.query.all()
|
||
|
|
matched = []
|
||
|
|
|
||
|
|
for room in rooms:
|
||
|
|
if not room.name:
|
||
|
|
continue
|
||
|
|
room_name_lower = room.name.lower()
|
||
|
|
# Match exact ou partiel
|
||
|
|
if room_name_lower in text_lower:
|
||
|
|
matched.append(room)
|
||
|
|
continue
|
||
|
|
# Match par code si la salle a un code
|
||
|
|
if hasattr(room, 'code') and room.code:
|
||
|
|
if room.code.lower() in text_lower:
|
||
|
|
matched.append(room)
|
||
|
|
continue
|
||
|
|
|
||
|
|
# Dedoublonner
|
||
|
|
seen_ids = set()
|
||
|
|
unique = []
|
||
|
|
for r in matched:
|
||
|
|
if r.id not in seen_ids:
|
||
|
|
unique.append(r)
|
||
|
|
seen_ids.add(r.id)
|
||
|
|
|
||
|
|
return unique
|
||
|
|
|
||
|
|
|
||
|
|
def extract_zone_from_text(text):
|
||
|
|
"""Cherche des zones mentionnees (cour, couloir, batiment A, etc.)."""
|
||
|
|
if not text:
|
||
|
|
return []
|
||
|
|
|
||
|
|
text_lower = text.lower()
|
||
|
|
buildings = Building.query.all()
|
||
|
|
matched = []
|
||
|
|
|
||
|
|
for b in buildings:
|
||
|
|
if not b.name:
|
||
|
|
continue
|
||
|
|
if b.name.lower() in text_lower:
|
||
|
|
matched.append(b)
|
||
|
|
|
||
|
|
# Patterns communs
|
||
|
|
common_zones = ['cour', 'couloir', 'gymnase', 'cantine', 'cdi', 'salle des professeurs',
|
||
|
|
'chaufferie', 'garage', 'atelier', 'preau']
|
||
|
|
for zone in common_zones:
|
||
|
|
if zone in text_lower:
|
||
|
|
matched.append(type('Zone', (), {'name': zone.capitalize(), 'id': None})())
|
||
|
|
|
||
|
|
return matched
|
||
|
|
|
||
|
|
|
||
|
|
def enrich_interpretation_with_location(text, suggested_location=None, suggested_equipment=None):
|
||
|
|
"""Complete les informations de localisation manquantes depuis le texte.
|
||
|
|
|
||
|
|
Si l'IA n'a pas trouve la salle, on cherche dans le texte.
|
||
|
|
"""
|
||
|
|
result = {
|
||
|
|
'rooms': [],
|
||
|
|
'zones': [],
|
||
|
|
'location': suggested_location or '',
|
||
|
|
}
|
||
|
|
|
||
|
|
# Si pas de location suggeree, chercher dans le texte
|
||
|
|
if not suggested_location:
|
||
|
|
rooms = extract_rooms_from_text(text)
|
||
|
|
if rooms:
|
||
|
|
result['rooms'] = [{'id': r.id, 'name': r.name} for r in rooms]
|
||
|
|
result['location'] = rooms[0].name
|
||
|
|
|
||
|
|
zones = extract_zone_from_text(text)
|
||
|
|
if zones:
|
||
|
|
result['zones'] = [z.name for z in zones]
|
||
|
|
if not result['location']:
|
||
|
|
result['location'] = zones[0].name
|
||
|
|
|
||
|
|
return result
|