nacim/Presidio
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Actualiser app.py

Browse Source
This commit is contained in:
nacim
2025-08-04 05:39:03 +00:00
parent adf495b792
commit 29e19fcd22
1 changed files with 1 additions and 23 deletions
Download Patch File Download Diff File Expand all files Collapse all files

24
app.py
View File

@@ -23,29 +23,7 @@ except Exception as e:
logger.exception("Error during AnalyzerEngine initialization.") logger.exception("Error during AnalyzerEngine initialization.")
analyzer = None analyzer = None
# Regex strict pour IBAN belge format attendu
IBAN_REGEX = re.compile(r"\b[A-Z]{2}[0-9]{2}(?:\s[0-9]{4}){3}\b", re.IGNORECASE)
# Regex IPv4
IPV4_REGEX = re.compile(
r"\b(?:(?:25[0-5]|2[0-4][0-9]|1\d{2}|[1-9]?\d)\.){3}"
r"(?:25[0-5]|2[0-4][0-9]|1\d{2}|[1-9]?\d)\b"
)
# Liste des labels/phrases à exclure d’anonymisation (en minuscules)
IGNORE_LABELS = {
"témoins",
"témoins clés",
"coordonnées",
"coordonnées bancaires",
"contexte financier",
"données sensibles",
"contexte",
"montrent",
"montrent des",
"montrent des irrégularités",
"bénéficiaire",
}
def normalize_label(text: str) -> str: def normalize_label(text: str) -> str:
return text.strip().lower() return text.strip().lower()
@@ -91,7 +69,7 @@ def analyze_text():
logger.warning(f"Invalid IBAN detected, skipping: '{ent_text}'") logger.warning(f"Invalid IBAN detected, skipping: '{ent_text}'")
continue continue
# Recadrage IP_ADDRESS strict IPv4 (wildcard possible pour IPv6 si besoin) # Recadrage IP_ADDRESS strict IPv4
if res.entity_type == "IP_ADDRESS": if res.entity_type == "IP_ADDRESS":
match = IPV4_REGEX.search(ent_text) match = IPV4_REGEX.search(ent_text)
if match: if match: