Dom/anonymisation
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
133 Commits 3 Branches 1 Tag
7fc97aa11f44c7aa85b490ce4163364404f7c657
Go to file
Clone
Open with VS Code Open with VSCodium Open with Intellij IDEA
Download ZIP Download TAR.GZ Download BUNDLE
Domi31tls 7fc97aa11f feat(q1): add quarantine.py module — entries, manager, logger 
Module standalone pour la quarantaine différentielle Q-1 :
- QuarantineEntry dataclass (doc_name, reason, detail, severity, flags...)
- QuarantineManager (flag, has_full_quarantine, finalize, INDEX.md gen)
- DocLogger (B-2 logs par doc, append-only)
- Constantes SEUIL_TEXTE_MINI=100, SEUIL_RESCAN_RESIDUEL=0

Smoke test OK : 2 entrées (full + partial), INDEX.md, errors.log,
reason.txt générés conformes spec §6 du consolidé v2.

Ref: docs/coordination/inbox/for-dom/2026-05-29_consolide_pseudocode-Q1-v2.md

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-29 17:58:46 +02:00
.github/workflows
Ajout build portable : Python embarqué + lanceur .bat (sans compilation)
2026-02-17 23:27:32 +01:00
.kiro/specs/anonymization-quality-optimization
docs(phase1): Résumé exécutif Phase 1 pour l'utilisateur
2026-03-02 23:37:42 +01:00
.snapshots
feat: Phase 1 - Système d'évaluation de la qualité
2026-03-02 10:07:41 +01:00
ano
chore: add .gitignore, remove PDFs/models/zips from history
2026-03-05 00:37:19 +01:00
assets
feat(ui): refonte UI — logo aivanonym + palette magenta/pêche + onglets + v5.5
2026-04-15 22:04:41 +02:00
config
feat(admin_rules): CLI simulate_admin_rule + fix email avant force_terms
2026-04-28 12:02:17 +02:00
corpus_validation
gui: Ajout indicateurs qualité (fuites, performances)
2026-03-02 21:34:18 +01:00
corpus_validation_sample
feat: Validation corpus complet - 100% qualité confirmée
2026-03-02 19:55:48 +01:00
data
Externalize dictionaries and add anonymization review corpus
2026-04-21 10:32:57 +02:00
detectors
feat: réduction FP + gazetteers adresses FINESS + batch parallèle + corrections multi-axes
2026-03-16 09:26:56 +01:00
docs
Add human review protocol and admin rules contract
2026-04-21 10:59:02 +02:00
evaluation
feat: réduction FP + gazetteers adresses FINESS + batch parallèle + corrections multi-axes
2026-03-16 09:26:56 +01:00
pdf_natif
chore: add .gitignore, remove PDFs/models/zips from history
2026-03-05 00:37:19 +01:00
regression_tests
feat(phase2): Multi-signal NER — BDPM gazetteers, confiance EDS, safe patterns, GLiNER
2026-03-09 12:01:46 +01:00
schemas
Add human review protocol and admin rules contract
2026-04-21 10:59:02 +02:00
scripts
Externalize dictionaries and add anonymization review corpus
2026-04-21 10:32:57 +02:00
test_3ogc
feat: Phase 1 - Système d'évaluation de la qualité
2026-03-02 10:07:41 +01:00
test_anonymise
feat: Phase 1 - Système d'évaluation de la qualité
2026-03-02 10:07:41 +01:00
test_chcb_leak
docs: Analyse finale validation corpus - système fonctionnel
2026-03-02 21:38:30 +01:00
test_doctr_fix
gui: Ajout indicateurs qualité (fuites, performances)
2026-03-02 21:34:18 +01:00
test_gui_output
fix(gui): Retirer paramètre use_vlm non supporté par process_pdf
2026-03-02 21:53:54 +01:00
tests
feat(admin_rules): CLI simulate_admin_rule + fix email avant force_terms
2026-04-28 12:02:17 +02:00
tools
feat(admin_rules): CLI simulate_admin_rule + fix email avant force_terms
2026-04-28 12:02:17 +02:00
.~lock.FONCTIONNEMENT.md#
feat: Phase 1 - Système d'évaluation de la qualité
2026-03-02 10:07:41 +01:00
.gitignore
feat(gui): afficher version + build date + commit dans titre et status bar
2026-04-15 18:40:58 +02:00
admin_rules.py
Wire admin rules into ONNX anonymizer
2026-04-21 12:10:17 +02:00
analyze_anonymization_result.py
demo: Test d'anonymisation sur document réel
2026-03-02 10:19:55 +01:00
anonymisation_onefile.spec
feat(ui): refonte UI — logo aivanonym + palette magenta/pêche + onglets + v5.5
2026-04-15 22:04:41 +02:00
anonymizer_core_refactored_onnx.py
feat(admin_rules): CLI simulate_admin_rule + fix email avant force_terms
2026-04-28 12:02:17 +02:00
anonymizer_core_refactored.py
Externalize dictionaries and add anonymization review corpus
2026-04-21 10:32:57 +02:00
audit_fp_detector.py
Stop words +170 : détection automatique FP via dictionnaire français (audit_fp_detector.py)
2026-02-28 10:04:33 +01:00
build_windows.bat
Externalize dictionaries and add anonymization review corpus
2026-04-21 10:32:57 +02:00
camembert_ner_manager.py
feat(phase3): CamemBERT v3 + détection villes + initiales + texte espacé + docs réglementaires
2026-03-11 12:16:13 +01:00
config_defaults.py
Externalize dictionaries and add anonymization review corpus
2026-04-21 10:32:57 +02:00
corpus_validation_full.log
gui: Ajout indicateurs qualité (fuites, performances)
2026-03-02 21:34:18 +01:00
corpus_validation.log
feat: Validation corpus complet - 100% qualité confirmée
2026-03-02 19:55:48 +01:00
demo_complete_anonymization.py
demo: Test d'anonymisation sur document réel
2026-03-02 10:19:55 +01:00
demo_evaluation.py
demo: Ajout script de démonstration et correction tests
2026-03-02 10:14:56 +01:00
eds_pseudo_manager.py
feat(phase2): Multi-signal NER — BDPM gazetteers, confiance EDS, safe patterns, GLiNER
2026-03-09 12:01:46 +01:00
FONCTIONNEMENT.md
Externalize dictionaries and add anonymization review corpus
2026-04-21 10:32:57 +02:00
format_converter.py
feat: support multi-formats — DOCX, images, ODT, RTF, TXT, HTML
2026-03-18 09:25:26 +01:00
gliner_manager.py
feat(phase2): Multi-signal NER — BDPM gazetteers, confiance EDS, safe patterns, GLiNER
2026-03-09 12:01:46 +01:00
install.sh
Initial commit — Pseudonymisation de PDF v5
2026-02-16 15:03:37 +01:00
launcher.py
fix(splash): étapes de chargement dans le splash NATIF (pas le tkinter)
2026-04-15 23:34:40 +02:00
ner_manager_onnx.py
Initial commit — Pseudonymisation de PDF v5
2026-02-16 15:03:37 +01:00
pdf_mask_designer.py
Initial commit — Pseudonymisation de PDF v5
2026-02-16 15:03:37 +01:00
Pseudonymisation_Gui_Models_V4.py
Externalize dictionaries and add anonymization review corpus
2026-04-21 10:32:57 +02:00
Pseudonymisation_Gui_V5.py
Externalize dictionaries and add anonymization review corpus
2026-04-21 10:32:57 +02:00
pseudonymisation_pipeline_gui_v3.py
Externalize dictionaries and add anonymization review corpus
2026-04-21 10:32:57 +02:00
Pseudonymisation_Pipeline_Robuste_Patch.py
Initial commit — Pseudonymisation de PDF v5
2026-02-16 15:03:37 +01:00
pseudonymisation_pipeline_robuste.py
Initial commit — Pseudonymisation de PDF v5
2026-02-16 15:03:37 +01:00
qc_audit.py
Amélioration qualité anonymisation : dico médicaments auto, noms soignants, garde trackare, validation EDS, QC audit
2026-02-26 18:58:18 +01:00
quarantine.py
feat(q1): add quarantine.py module — entries, manager, logger
2026-05-29 17:58:46 +02:00
readme.md
Initial commit — Pseudonymisation de PDF v5
2026-02-16 15:03:37 +01:00
requirements.txt
chore(deps): rendre python-doctr requis (OCR systématique)
2026-04-27 16:17:41 +02:00
run_batch_30_audit.py
Externalize dictionaries and add anonymization review corpus
2026-04-21 10:32:57 +02:00
run_batch_59ogc.py
Externalize dictionaries and add anonymization review corpus
2026-04-21 10:32:57 +02:00
run_batch_silver_export.py
Externalize dictionaries and add anonymization review corpus
2026-04-21 10:32:57 +02:00
server.py
Externalize dictionaries and add anonymization review corpus
2026-04-21 10:32:57 +02:00
setup_env_and_build.bat
Initial commit — Pseudonymisation de PDF v5
2026-02-16 15:03:37 +01:00
test_anonymization_example.py
demo: Test d'anonymisation sur document réel
2026-03-02 10:19:55 +01:00
test_gui_error.py
Externalize dictionaries and add anonymization review corpus
2026-04-21 10:32:57 +02:00
test_gui_fixed.py
Externalize dictionaries and add anonymization review corpus
2026-04-21 10:32:57 +02:00
test-mini.js
chore: add .gitignore, remove PDFs/models/zips from history
2026-03-05 00:37:19 +01:00
vlm_manager.py
Sécurité VLM : format JSON forcé, modèle local uniquement, fix logging critique
2026-02-27 02:38:30 +01:00

readme.md

placer tout les fichiers dans un répertoire. faire un chmod 777 install.sh pour lui donner les droits d'execution lancer ./install.sh pour lancer l'installation complete

L'installation peut prendre du temps, elle charge deux modele IA nlp. Elle crée un environement virtuel python.

Reference in New Issue View Git Blame Copy Permalink
Description
No description provided
Readme 283 MiB
Languages
Python 98.2%
Batchfile 1%
PowerShell 0.5%
Shell 0.3%