theo/extracteur-pdf
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Rendre le script portable (Docker/Windows) et ajout de l'auto-ouverture du navigateur

Browse Source
This commit is contained in:
theo
2026-04-15 18:44:44 +02:00
parent 5b9c163670
commit 2d6d747371
1 changed files with 25 additions and 11 deletions
Download Patch File Download Diff File Expand all files Collapse all files
webapp.py
+25 -11
View File
@@ -6,18 +6,24 @@ import pandas as pd
import pdfplumber
import unicodedata
import traceback
import webbrowser
from threading import Timer
from flask import Flask, request, render_template_string, Response, send_from_directory, redirect
app = Flask(__name__)
# Dossiers et fichiers
TEMP_DIR = "/app/downloads"
if not os.path.exists(TEMP_DIR): os.makedirs(TEMP_DIR)
# --- ADAPTATION DES CHEMINS (AUTO-DETECTION) ---
# Si on est dans Docker, on utilise /app, sinon le dossier local
BASE_DIR = "/app" if os.path.exists("/app") else os.getcwd()
TEMP_DIR = os.path.join(BASE_DIR, "downloads")
if not os.path.exists(TEMP_DIR):
os.makedirs(TEMP_DIR)
TEMP_DATA_FILE = os.path.join(TEMP_DIR, "temp_data.json")
TEMP_MISSING_FILE = os.path.join(TEMP_DIR, "temp_missing.json")
CSV_FILE = "/app/corresp_IGPDE.csv"
ODS_FILE = "/app/corresp_IGPDE.ods"
CSV_FILE = os.path.join(BASE_DIR, "corresp_IGPDE.csv")
ODS_FILE = os.path.join(BASE_DIR, "corresp_IGPDE.ods")
RESULT_FILE = os.path.join(TEMP_DIR, "resultat.xlsx")
# --- GESTION DE LA BASE DE DONNÉES ---
@@ -47,11 +53,10 @@ COMMON_STYLE = """
.nav-btn { background: #fff; border: 1px solid #cbd5e0; padding: 8px 15px; border-radius: 8px; text-decoration: none; color: #4a5568; font-size: 0.9rem; transition: 0.2s; }
.nav-btn:hover { background: #edf2f7; }
.nav-btn.active { background: #3182ce; color: white; border-color: #3182ce; }
.card { background: white; padding: 2rem; border-radius: 15px; box-shadow: 0 4px 15px rgba(0,0,0,0.05); max-width: 1200px; margin: 0 auto; }
.card { background: white; padding: 2rem; border-radius: 15px; box-shadow: 0 4px 15px rgba(0,0,0,0.05); max-width: 1100px; margin: 0 auto; }
h1 { color: #2d3748; margin-top: 0; font-size: 1.5rem; }
.header-box { display: flex; justify-content: space-between; align-items: center; margin-bottom: 20px; }
.btn-submit { background: #3182ce; color: white; border: none; padding: 10px 20px; border-radius: 8px; font-size: 0.9rem; cursor: pointer; font-weight: 600; text-decoration: none; display: inline-block; }
.btn-submit:hover { background: #2b6cb0; }
#log-window { background: #1a202c; color: #cbd5e0; padding: 15px; border-radius: 10px; height: 350px; overflow-y: auto; font-family: 'Courier New', monospace; font-size: 0.85rem; text-align: left; margin-top: 20px; line-height: 1.6; border: 1px solid #2d3748; }
.log-success { color: #68d391; font-weight: bold; }
.log-warn { color: #fc8181; }
@@ -129,6 +134,7 @@ def process():
data[col] = row[1].replace('\n', ' ').strip()
break
all_formations.append(data)
# Nettoyage sécurisé pour les logs JS
t_log = titre_brut[:40].replace("'", " ").replace('"', ' ')
yield f"<script>addLog('Page {page_num} : {t_log}...', 'success', {percent})</script>"
except Exception: pass
@@ -194,14 +200,22 @@ def finalize():
f['Nouvelle nomenclature'] = mapping_dict.get(super_norm(key), mapping_dict.get(super_norm(f['Chapitre']), ""))
df = pd.DataFrame(all_formations).fillna('')
cols = ['Nouvelle nomenclature', 'Chapitre', 'Sous-chapitre', 'Titre', 'Page_Source'] + [c for c in df.columns if c not in ['Nouvelle nomenclature', 'Chapitre', 'Sous-chapitre', 'Titre', 'Page_Source']]
df_f = df[[c for c in cols if c in df.columns]]
df_f.to_excel(RESULT_FILE, index=False, engine='xlsxwriter')
df_cols = [c for c in cols if c in df.columns]
df[df_cols].to_excel(RESULT_FILE, index=False, engine='xlsxwriter')
return redirect("/view_result")
@app.route("/download")
def download(): return send_from_directory(TEMP_DIR, "resultat.xlsx", as_attachment=True)
@app.route("/download_db")
def download_db(): return send_from_directory("/app", "corresp_IGPDE.csv", as_attachment=True)
def download_db(): return send_from_directory(BASE_DIR, "corresp_IGPDE.csv", as_attachment=True)
if __name__ == "__main__": app.run(host='0.0.0.0', port=5000)
# --- LANCEMENT AUTOMATIQUE ---
def open_browser():
webbrowser.open_new("http://127.0.0.1:5000")
if __name__ == "__main__":
# N'ouvre le navigateur que si on n'est PAS dans Docker
if not os.path.exists("/app"):
Timer(1.5, open_browser).start()
app.run(host='0.0.0.0', port=5000)