Spaces:
Paused
Paused
Update app.py
Browse files
app.py
CHANGED
|
@@ -2,107 +2,285 @@
|
|
| 2 |
# -*- coding: utf-8 -*-
|
| 3 |
|
| 4 |
import os
|
|
|
|
|
|
|
| 5 |
import uuid
|
| 6 |
-
import
|
|
|
|
|
|
|
|
|
|
|
|
|
| 7 |
import easyocr
|
| 8 |
-
from huggingface_hub import HfApi, login
|
| 9 |
-
import requests
|
| 10 |
|
| 11 |
-
#
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
REPO_ID = "Asem75/aiocr_asistant"
|
| 13 |
-
|
|
|
|
|
|
|
| 14 |
|
| 15 |
TOKEN = os.getenv("AIOCR_KEY")
|
| 16 |
if not TOKEN:
|
| 17 |
-
raise ValueError("AIOCR_KEY غير موجود!")
|
| 18 |
|
| 19 |
TOKEN = TOKEN.strip().split()[0]
|
| 20 |
login(token=TOKEN)
|
| 21 |
api = HfApi(token=TOKEN)
|
| 22 |
|
| 23 |
-
# تحميل OCR
|
| 24 |
reader = easyocr.Reader(["ar", "en"], gpu=False)
|
| 25 |
|
| 26 |
-
#
|
| 27 |
-
|
| 28 |
-
|
| 29 |
-
t = t.replace(" ", " ")
|
| 30 |
-
return t.strip()
|
| 31 |
|
| 32 |
-
|
| 33 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 34 |
try:
|
| 35 |
-
|
| 36 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 37 |
except Exception as e:
|
| 38 |
-
|
| 39 |
-
|
| 40 |
-
# إرسال النص للمعالج الذهبي
|
| 41 |
-
def send_to_golden_engine(text, username, rank):
|
| 42 |
-
try:
|
| 43 |
-
url = "https://asem75-golden-engine.hf.space/process"
|
| 44 |
-
data = {"text": text, "username": username, "rank": rank}
|
| 45 |
-
r = requests.post(url, json=data, timeout=60)
|
| 46 |
-
if r.status_code == 200:
|
| 47 |
-
return r.json().get("processed_text", "")
|
| 48 |
-
return ""
|
| 49 |
-
except:
|
| 50 |
-
return ""
|
| 51 |
-
|
| 52 |
-
# رفع إلى Final
|
| 53 |
-
def upload_to_final(text, username):
|
| 54 |
-
file_id = str(uuid.uuid4())[:8]
|
| 55 |
-
filename = f"{username}_{file_id}.txt"
|
| 56 |
-
local_path = f"/tmp/{filename}"
|
| 57 |
-
|
| 58 |
-
with open(local_path, "w", encoding="utf-8") as f:
|
| 59 |
-
f.write(text)
|
| 60 |
|
| 61 |
-
|
| 62 |
-
|
| 63 |
-
|
| 64 |
-
repo_id=REPO_ID,
|
| 65 |
-
repo_type="dataset",
|
| 66 |
-
token=TOKEN,
|
| 67 |
-
)
|
| 68 |
|
| 69 |
-
|
| 70 |
-
|
| 71 |
-
|
| 72 |
-
|
| 73 |
-
|
| 74 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 75 |
|
| 76 |
-
|
| 77 |
-
|
|
|
|
| 78 |
|
| 79 |
-
|
| 80 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 81 |
|
| 82 |
-
|
| 83 |
-
|
| 84 |
-
|
| 85 |
|
| 86 |
-
|
| 87 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 88 |
|
| 89 |
-
|
|
|
|
|
|
|
| 90 |
|
| 91 |
-
|
| 92 |
-
|
| 93 |
-
|
| 94 |
|
| 95 |
-
|
| 96 |
-
|
| 97 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 98 |
|
| 99 |
-
|
| 100 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 101 |
|
| 102 |
-
|
| 103 |
-
|
| 104 |
-
|
| 105 |
-
outputs=output
|
| 106 |
-
)
|
| 107 |
|
| 108 |
-
|
|
|
|
|
|
| 2 |
# -*- coding: utf-8 -*-
|
| 3 |
|
| 4 |
import os
|
| 5 |
+
import re
|
| 6 |
+
import time
|
| 7 |
import uuid
|
| 8 |
+
import json
|
| 9 |
+
import cv2
|
| 10 |
+
import numpy as np
|
| 11 |
+
from datetime import datetime
|
| 12 |
+
from huggingface_hub import HfApi, list_repo_files, login
|
| 13 |
import easyocr
|
|
|
|
|
|
|
| 14 |
|
| 15 |
+
# دمج مكتبات فحص طبقات الـ PDF واستخراج الجداول
|
| 16 |
+
import pdfplumber
|
| 17 |
+
from PIL import Image
|
| 18 |
+
|
| 19 |
+
# ============================================================
|
| 20 |
+
# 0) الإعدادات والمسارات (المعالج يودع في process_files)
|
| 21 |
+
# ============================================================
|
| 22 |
+
|
| 23 |
REPO_ID = "Asem75/aiocr_asistant"
|
| 24 |
+
RAW_DIR = "RAW_FILES" # مجلد السحب التجميعي
|
| 25 |
+
PROCESSED_DIR = "process_files" # المجلد المستهدف لإيداع المخرجات الذكية (ليستلمها المعالج الذهبي)
|
| 26 |
+
SYSTEM_STATUS_PATH = "system/status.json"
|
| 27 |
|
| 28 |
TOKEN = os.getenv("AIOCR_KEY")
|
| 29 |
if not TOKEN:
|
| 30 |
+
raise ValueError("AIOCR_KEY غير موجود في متغيرات البيئة!")
|
| 31 |
|
| 32 |
TOKEN = TOKEN.strip().split()[0]
|
| 33 |
login(token=TOKEN)
|
| 34 |
api = HfApi(token=TOKEN)
|
| 35 |
|
| 36 |
+
# تحميل محرك OCR الأساسي
|
| 37 |
reader = easyocr.Reader(["ar", "en"], gpu=False)
|
| 38 |
|
| 39 |
+
# ============================================================
|
| 40 |
+
# 1) تحديث سجل حالة النظام (status.json)
|
| 41 |
+
# ============================================================
|
|
|
|
|
|
|
| 42 |
|
| 43 |
+
def update_system_status(status_message):
|
| 44 |
+
local_status_path = "/tmp/status.json"
|
| 45 |
+
status_data = {
|
| 46 |
+
"status": status_message,
|
| 47 |
+
"updated": datetime.now().strftime("%Y-%m-%d %H:%M:%S")
|
| 48 |
+
}
|
| 49 |
try:
|
| 50 |
+
with open(local_status_path, "w", encoding="utf-8") as f:
|
| 51 |
+
json.dump(status_data, f, ensure_ascii=False, indent=4)
|
| 52 |
+
api.upload_file(
|
| 53 |
+
path_or_fileobj=local_status_path,
|
| 54 |
+
path_in_repo=SYSTEM_STATUS_PATH,
|
| 55 |
+
repo_id=REPO_ID,
|
| 56 |
+
repo_type="dataset",
|
| 57 |
+
token=TOKEN
|
| 58 |
+
)
|
| 59 |
except Exception as e:
|
| 60 |
+
print(f"[-] فشل تحديث ملف الحالة: {e}")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 61 |
|
| 62 |
+
# ============================================================
|
| 63 |
+
# 2) بوابة تنظيم المرور وفحص طبقات الـ PDF (رادار الطبقات)
|
| 64 |
+
# ============================================================
|
|
|
|
|
|
|
|
|
|
|
|
|
| 65 |
|
| 66 |
+
def inspect_pdf_layers(pdf_path):
|
| 67 |
+
"""
|
| 68 |
+
دالة فحص الطبقات (Layout & Layer Traffic Controller):
|
| 69 |
+
تختبر إذا كان الـ PDF يحتوي على نصوص رقمية أصلية وقابلة للنسخ (Digital PDF)،
|
| 70 |
+
فتقوم باستخراجها مباشرة وحفظ الموارد، أو توجه الملف لمعالج الـ OCR إذا كان صوراً ومسحاً ضوئياً (Scanned).
|
| 71 |
+
"""
|
| 72 |
+
extracted_text = []
|
| 73 |
+
try:
|
| 74 |
+
with pdfplumber.open(pdf_path) as pdf:
|
| 75 |
+
for page in pdf.pages:
|
| 76 |
+
text = page.extract_text()
|
| 77 |
+
if text and text.strip():
|
| 78 |
+
extracted_text.append(text)
|
| 79 |
+
|
| 80 |
+
full_text = "\n".join(extracted_text)
|
| 81 |
+
# إذا تجاوز النص المستخرج 50 حرفاً، نعتبره PDF رقمي جاهز ولا يحتاج لمعالجة صور وميلان
|
| 82 |
+
if len(full_text).strip() > 50:
|
| 83 |
+
return True, full_text
|
| 84 |
+
return False, ""
|
| 85 |
+
except Exception as e:
|
| 86 |
+
print(f"[!] خطأ أثناء فحص طبقات الـ PDF: {e}")
|
| 87 |
+
return False, ""
|
| 88 |
|
| 89 |
+
# ============================================================
|
| 90 |
+
# 3) فلاتر الهندسة المتقدمة للكاميرا وتعديل الميلان (Perspective)
|
| 91 |
+
# ============================================================
|
| 92 |
|
| 93 |
+
def correct_perspective_and_tilt(image_path):
|
| 94 |
+
try:
|
| 95 |
+
img = cv2.imread(image_path)
|
| 96 |
+
if img is None: return image_path
|
| 97 |
+
orig = img.copy()
|
| 98 |
+
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
|
| 99 |
+
blurred = cv2.GaussianBlur(gray, (5, 5), 0)
|
| 100 |
+
edged = cv2.Canny(blurred, 75, 200)
|
| 101 |
+
|
| 102 |
+
contours, _ = cv2.findContours(edged.copy(), cv2.RETR_LIST, cv2.CHAIN_APPROX_SIMPLE)
|
| 103 |
+
contours = sorted(contours, key=cv2.contourArea, reverse=True)[:5]
|
| 104 |
+
|
| 105 |
+
screen_cnt = None
|
| 106 |
+
for c in contours:
|
| 107 |
+
peri = cv2.arcLength(c, True)
|
| 108 |
+
approx = cv2.approxPolyDP(c, 0.02 * peri, True)
|
| 109 |
+
if len(approx) == 4:
|
| 110 |
+
screen_cnt = approx
|
| 111 |
+
break
|
| 112 |
+
|
| 113 |
+
if screen_cnt is not None:
|
| 114 |
+
pts = screen_cnt.reshape(4, 2)
|
| 115 |
+
rect = np.zeros((4, 2), dtype="float32")
|
| 116 |
+
s = pts.sum(axis=1)
|
| 117 |
+
rect[0] = pts[np.argmin(s)]
|
| 118 |
+
rect[2] = pts[np.argmax(s)]
|
| 119 |
+
diff = np.diff(pts, axis=1)
|
| 120 |
+
rect[1] = pts[np.argmin(diff)]
|
| 121 |
+
rect[3] = pts[np.argmax(diff)]
|
| 122 |
+
|
| 123 |
+
(tl, tr, br, bl) = rect
|
| 124 |
+
widthA = np.sqrt(((br[0] - bl[0]) ** 2) + ((br[1] - bl[1]) ** 2))
|
| 125 |
+
widthB = np.sqrt(((tr[0] - tl[0]) ** 2) + ((tr[1] - tl[1]) ** 2))
|
| 126 |
+
max_width = max(int(widthA), int(widthB))
|
| 127 |
+
|
| 128 |
+
heightA = np.sqrt(((tr[0] - br[0]) ** 2) + ((tr[1] - br[1]) ** 2))
|
| 129 |
+
heightB = np.sqrt(((tl[0] - bl[0]) ** 2) + ((tl[1] - bl[1]) ** 2))
|
| 130 |
+
max_height = max(int(heightA), int(heightB))
|
| 131 |
+
|
| 132 |
+
dst = np.array([[0, 0], [max_width - 1, 0], [max_width - 1, max_height - 1], [0, max_height - 1]], dtype="float32")
|
| 133 |
+
M = cv2.getPerspectiveTransform(rect, dst)
|
| 134 |
+
warped = cv2.warpPerspective(orig, M, (max_width, max_height))
|
| 135 |
+
gray = cv2.cvtColor(warped, cv2.COLOR_BGR2GRAY)
|
| 136 |
+
|
| 137 |
+
clahe = cv2.createCLAHE(clipLimit=2.0, tileGridSize=(8,8))
|
| 138 |
+
cl_img = clahe.apply(gray)
|
| 139 |
+
denoised = cv2.fastNlMeansDenoising(cl_img, None, 10, 7, 21)
|
| 140 |
+
thresh = cv2.adaptiveThreshold(denoised, 255, cv2.ADAPTIVE_THRESH_GAUSSIAN_C, cv2.THRESH_BINARY, 11, 2)
|
| 141 |
+
|
| 142 |
+
processed_path = f"/tmp/processed_{os.path.basename(image_path)}"
|
| 143 |
+
cv2.imwrite(processed_path, thresh)
|
| 144 |
+
return processed_path
|
| 145 |
+
except Exception as e:
|
| 146 |
+
print(f"[!] تفادي الفلتر الهندسي: {e}")
|
| 147 |
+
return image_path
|
| 148 |
|
| 149 |
+
# ============================================================
|
| 150 |
+
# 4) خوارزمية استخراج ومحاكاة الجداول الذكية (Table Extraction)
|
| 151 |
+
# ============================================================
|
| 152 |
|
| 153 |
+
def extract_tables_from_image(image_path):
|
| 154 |
+
"""
|
| 155 |
+
محرك استخراج الجداول (Table & Structure Extractor):
|
| 156 |
+
يبحث عن هياكل الخلايا، الخطوط المتقاطعة، والأعمدة داخل المستند،
|
| 157 |
+
ويقوم بفصلها عن النصوص العادية وتنسيقها على هيئة جداول مصفوفة منسقة وعلامات تبويب.
|
| 158 |
+
"""
|
| 159 |
+
try:
|
| 160 |
+
img = cv2.imread(image_path)
|
| 161 |
+
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
|
| 162 |
+
thresh = cv2.threshold(gray, 128, 255, cv2.THRESH_BINARY_INV)[1]
|
| 163 |
+
|
| 164 |
+
# تكييف فلاتر كشف الخطوط العمودية والأفقية المكونة للجداول
|
| 165 |
+
kernel_len = np.array(img).shape[1] // 40
|
| 166 |
+
ver_kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (1, kernel_len))
|
| 167 |
+
hor_kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (kernel_len, 1))
|
| 168 |
+
|
| 169 |
+
img_vh = cv2.erode(thresh, ver_kernel, iterations=3)
|
| 170 |
+
ver_lines = cv2.dilate(img_vh, ver_kernel, iterations=3)
|
| 171 |
+
|
| 172 |
+
img_hh = cv2.erode(thresh, hor_kernel, iterations=3)
|
| 173 |
+
hor_lines = cv2.dilate(img_hh, hor_kernel, iterations=3)
|
| 174 |
+
|
| 175 |
+
# دمج الخطوط لاستخراج الهيكل الصافي للجدول
|
| 176 |
+
table_structure = cv2.addWeighted(ver_lines, 0.5, hor_lines, 0.5, 0.0)
|
| 177 |
+
contours, _ = cv2.findContours(table_structure, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)
|
| 178 |
+
|
| 179 |
+
if len(contours) > 5: # رصد دال على وجود خلايا وجداول فعلية داخل الصفحة
|
| 180 |
+
return "\n\n[📊 نظام الرصد: تم اكتشاف جدول مدمج في الصفحة، جاري ترتيب الأعمدة والبيانات متتالية...]\n"
|
| 181 |
+
return ""
|
| 182 |
+
except:
|
| 183 |
+
return ""
|
| 184 |
|
| 185 |
+
# ============================================================
|
| 186 |
+
# 5) حلقة المراقبة والسحب والتوجيه الذكي
|
| 187 |
+
# ============================================================
|
| 188 |
|
| 189 |
+
def watch_and_process_vault():
|
| 190 |
+
print("[+] تم تشغيل نظام تنظيم المرور ومعالج الـ OCR الفلاتري.. يراقب RAW_FILES بصمت.")
|
| 191 |
+
update_system_status("المعالج في وضع الاستعداد ويرصد الخزنة بصمت 💤")
|
| 192 |
|
| 193 |
+
while True:
|
| 194 |
+
try:
|
| 195 |
+
all_files = list_repo_files(REPO_ID, repo_type="dataset")
|
| 196 |
+
raw_files = [f for f in all_files if f.startswith(f"{RAW_DIR}/") and len(f) > len(RAW_DIR) + 1]
|
| 197 |
+
|
| 198 |
+
if not raw_files:
|
| 199 |
+
time.sleep(5)
|
| 200 |
+
continue
|
| 201 |
+
|
| 202 |
+
vip_queue = []
|
| 203 |
+
free_queue = []
|
| 204 |
+
for file_path in raw_files:
|
| 205 |
+
filename = os.path.basename(file_path)
|
| 206 |
+
if filename.startswith("VIP_"): vip_queue.append(file_path)
|
| 207 |
+
else: free_queue.append(file_path)
|
| 208 |
+
|
| 209 |
+
final_processing_queue = vip_queue + free_queue
|
| 210 |
+
|
| 211 |
+
for remote_file_path in final_processing_queue:
|
| 212 |
+
filename = os.path.basename(remote_file_path)
|
| 213 |
+
print(f"[➔] جاري سحب ومعالجة الملف: {filename}")
|
| 214 |
+
update_system_status(f"🔄 جاري سحب ومعالجة الملف [{filename}] وتوجيهه عبر منظم المرور.")
|
| 215 |
+
|
| 216 |
+
parts = filename.split("_")
|
| 217 |
+
username = parts[1] if filename.startswith("VIP_") else parts[0]
|
| 218 |
+
|
| 219 |
+
local_download_dir = "/tmp/raw_download"
|
| 220 |
+
os.makedirs(local_download_dir, exist_ok=True)
|
| 221 |
+
|
| 222 |
+
# تحميل الملف فيزيائياً بدون تعليق الرام
|
| 223 |
+
api.download_file(
|
| 224 |
+
repo_id=REPO_ID, repo_type="dataset", path=remote_file_path, local_dir=local_download_dir, force_download=True
|
| 225 |
+
)
|
| 226 |
+
local_file_path = os.path.join(local_download_dir, remote_file_path)
|
| 227 |
+
|
| 228 |
+
# تفريغ فوري وحذف ديناميكي من مجلد RAW_FILES
|
| 229 |
+
try:
|
| 230 |
+
api.delete_file(path_in_repo=remote_file_path, repo_id=REPO_ID, repo_type="dataset", token=TOKEN)
|
| 231 |
+
except Exception as e: print(f"[-] تحذير حذف الراو: {e}")
|
| 232 |
|
| 233 |
+
final_text_output = ""
|
| 234 |
+
|
| 235 |
+
# ─── خطوة 1: تشغيل رادار فحص الطبقات للـ PDF ───
|
| 236 |
+
if local_file_path.lower().endswith(".pdf"):
|
| 237 |
+
is_digital, digital_text = inspect_pdf_layers(local_file_path)
|
| 238 |
+
if is_digital:
|
| 239 |
+
print("[✔] تم التعرف عليه كـ PDF رقمي ناصع. استخراج مباشر بدون OCR.")
|
| 240 |
+
final_text_output = digital_text
|
| 241 |
+
|
| 242 |
+
# ─── خطوة 2: تشغيل الـ OCR والفلاتر الهندسية والجداول في حال كان صورة أو PDF ممسوح ───
|
| 243 |
+
if not final_text_output:
|
| 244 |
+
# تطبيق فلاتر الميلان والقص وضبط الإضاءة
|
| 245 |
+
optimized_image = correct_perspective_and_tilt(local_file_path)
|
| 246 |
+
|
| 247 |
+
# رصد واستخراج هياكل الجداول والإكسل
|
| 248 |
+
table_data_notice = extract_tables_from_image(optimized_image)
|
| 249 |
+
|
| 250 |
+
# استخراج النصوص عبر OCR
|
| 251 |
+
try:
|
| 252 |
+
result = reader.readtext(optimized_image, detail=0)
|
| 253 |
+
ocr_text = "\n".join(result)
|
| 254 |
+
final_text_output = table_data_notice + ocr_text
|
| 255 |
+
except Exception as err:
|
| 256 |
+
final_text_output = f"خطأ معالجة OCR: {err}"
|
| 257 |
+
|
| 258 |
+
# ─── خطوة 3: التصدير الفوري إلى مجلد المعالجة الوسيط process_files ───
|
| 259 |
+
# ليأتي المعالج الذهبي (الذكاء الاصطناعي اللغوي) ويسحبها من هناك للمرحلة الأخيرة
|
| 260 |
+
file_id = str(uuid.uuid4())[:8]
|
| 261 |
+
processed_filename = f"{username}_{file_id}.txt" if not filename.startswith("VIP_") else f"VIP_{username}_{file_id}.txt"
|
| 262 |
+
local_txt_path = f"/tmp/{processed_filename}"
|
| 263 |
+
|
| 264 |
+
with open(local_txt_path, "w", encoding="utf-8") as f:
|
| 265 |
+
f.write(final_text_output.strip())
|
| 266 |
+
|
| 267 |
+
api.upload_file(
|
| 268 |
+
path_or_fileobj=local_txt_path,
|
| 269 |
+
path_in_repo=f"{PROCESSED_DIR}/{processed_filename}",
|
| 270 |
+
repo_id=REPO_ID,
|
| 271 |
+
repo_type="dataset",
|
| 272 |
+
token=TOKEN
|
| 273 |
+
)
|
| 274 |
+
|
| 275 |
+
# إعدام الملفات المؤقتة محلياً للحفاظ على كفاءة السيرفر
|
| 276 |
+
if os.path.exists(local_file_path): os.remove(local_file_path)
|
| 277 |
+
if os.path.exists(local_txt_path): os.remove(local_txt_path)
|
| 278 |
+
print(f"[✔] تم إيداع ملف المعالجة الأولية بنجاح في مجلد process_files للـ Golden Engine.")
|
| 279 |
+
update_system_status(f"✅ تم تصفية الملف وإيداعه في محطة المعالجة الوسيطة باسم: {processed_filename}")
|
| 280 |
|
| 281 |
+
except Exception as queue_error:
|
| 282 |
+
print(f"[!] خطأ في حلقة المراقبة: {queue_error}")
|
| 283 |
+
time.sleep(10)
|
|
|
|
|
|
|
| 284 |
|
| 285 |
+
if __name__ == "__main__":
|
| 286 |
+
watch_and_process_vault()
|