Update app.py
Browse files
app.py
CHANGED
@@ -2,22 +2,24 @@
|
|
2 |
import os
|
3 |
import re
|
4 |
import json
|
5 |
-
import time
|
6 |
import io
|
|
|
|
|
7 |
import fitz # PyMuPDF
|
8 |
import requests
|
9 |
import google.generativeai as genai
|
10 |
import mysql.connector
|
|
|
11 |
from bs4 import BeautifulSoup
|
12 |
from dotenv import load_dotenv
|
13 |
from tqdm import tqdm
|
14 |
from sentence_transformers import SentenceTransformer
|
15 |
import gradio as gr
|
16 |
-
from urllib.parse import urljoin, urlparse
|
17 |
from collections import deque
|
18 |
-
import pandas as pd
|
19 |
|
20 |
-
# ---
|
21 |
try:
|
22 |
import pytesseract
|
23 |
from PIL import Image
|
@@ -35,9 +37,8 @@ except ImportError:
|
|
35 |
DOCX_ENABLED = False
|
36 |
print("⚠️ تحذير: مكتبة python-docx غير موجودة. لن يتم قراءة ملفات وورد.")
|
37 |
|
38 |
-
|
39 |
# ============================================================
|
40 |
-
# 1.
|
41 |
# ============================================================
|
42 |
print("⏳ جاري تحميل الإعدادات...")
|
43 |
if os.path.exists('.env'):
|
@@ -48,8 +49,21 @@ DB_CONFIG = {
|
|
48 |
"user": os.getenv("DB_USER"),
|
49 |
"password": os.getenv("DB_PASSWORD"),
|
50 |
"database": os.getenv("DB_NAME"),
|
|
|
51 |
}
|
52 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
53 |
GOOGLE_API_KEY = os.getenv("GOOGLE_API_KEY")
|
54 |
if GOOGLE_API_KEY:
|
55 |
try:
|
@@ -63,103 +77,100 @@ else:
|
|
63 |
|
64 |
SCRAPE_URL = os.getenv("SCRAPE_URL", "https://fra.gov.eg/")
|
65 |
REQUEST_TIMEOUT = 60
|
66 |
-
PAGE_LIMIT =
|
|
|
67 |
|
68 |
-
print("⏳ جاري تحميل نموذج Embedding المحلي... (قد يستغرق بعض الوقت أول مرة)")
|
69 |
LOCAL_EMBEDDER = SentenceTransformer("all-MiniLM-L6-v2")
|
70 |
print("✅ تم تحميل نموذج Embedding المحلي.")
|
71 |
|
72 |
# ============================================================
|
73 |
-
# 2.
|
74 |
# ============================================================
|
75 |
def setup_database():
|
|
|
76 |
try:
|
77 |
-
|
78 |
-
|
79 |
-
|
80 |
-
|
81 |
-
|
82 |
-
conn = mysql.connector.connect(**DB_CONFIG)
|
83 |
-
cursor = conn.cursor()
|
84 |
-
print("✅ تم الاتصال بقاعدة البيانات بنجاح.")
|
85 |
-
|
86 |
-
cursor.execute("""
|
87 |
-
CREATE TABLE IF NOT EXISTS documents (
|
88 |
-
id INT AUTO_INCREMENT PRIMARY KEY,
|
89 |
-
source VARCHAR(2048) CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci NOT NULL UNIQUE,
|
90 |
-
content LONGTEXT CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci,
|
91 |
-
summary TEXT CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci,
|
92 |
-
status VARCHAR(50) DEFAULT 'pending',
|
93 |
-
created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
|
94 |
-
updated_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP,
|
95 |
-
filename VARCHAR(255) CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci NULL
|
96 |
-
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci;
|
97 |
-
""")
|
98 |
-
|
99 |
-
cursor.execute("""
|
100 |
-
CREATE TABLE IF NOT EXISTS document_chunks (
|
101 |
-
id INT AUTO_INCREMENT PRIMARY KEY,
|
102 |
-
document_id INT NOT NULL,
|
103 |
-
content TEXT CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci,
|
104 |
-
embedding JSON,
|
105 |
-
created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
|
106 |
-
updated_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP,
|
107 |
-
FOREIGN KEY (document_id) REFERENCES documents(id) ON DELETE CASCADE
|
108 |
-
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci;
|
109 |
-
""")
|
110 |
-
print(" -> جداول قاعدة البيانات جاهزة.")
|
111 |
-
conn.commit()
|
112 |
-
cursor.close()
|
113 |
-
conn.close()
|
114 |
return True, "قاعدة البيانات جاهزة."
|
115 |
except mysql.connector.Error as err:
|
116 |
-
error_msg = f"خطأ فادح في قاعدة البيانات: {err}"
|
117 |
-
print(
|
118 |
-
return False,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
119 |
|
120 |
# ============================================================
|
121 |
-
# 3.
|
122 |
# ============================================================
|
123 |
-
def crawl_site(start_url, page_limit=
|
124 |
visited = set([start_url])
|
125 |
queue = deque([start_url])
|
126 |
all_links = set([start_url])
|
127 |
base_domain = urlparse(start_url).netloc
|
128 |
-
|
129 |
with tqdm(total=page_limit, desc="🔍 Crawling Progress", unit="page") as pbar:
|
130 |
while queue and len(visited) < page_limit:
|
131 |
url = queue.popleft()
|
132 |
pbar.set_description(f"🔍 Crawling: {url[:70]}...")
|
133 |
-
|
134 |
try:
|
135 |
-
|
136 |
-
|
137 |
-
content_type =
|
138 |
-
|
139 |
if "text/html" in content_type:
|
140 |
-
|
141 |
-
|
|
|
142 |
for a_tag in soup.find_all("a", href=True):
|
143 |
href = urljoin(url, a_tag["href"]).split("#")[0].strip()
|
144 |
if not href.startswith("http") or urlparse(href).netloc != base_domain or href in visited:
|
145 |
continue
|
146 |
if len(visited) >= page_limit: break
|
147 |
-
|
148 |
visited.add(href)
|
149 |
all_links.add(href)
|
150 |
pbar.update(1)
|
151 |
-
# لا نضيف ملفات للزحف مرة أخرى
|
152 |
if not any(href.lower().endswith(ext) for ext in ['.pdf', '.docx', '.txt']):
|
153 |
queue.append(href)
|
154 |
-
except
|
155 |
-
tqdm.write(f"⚠️ خطأ أثناء الزحف إلى {url}: {e}")
|
156 |
continue
|
157 |
return list(all_links)
|
158 |
|
159 |
# ============================================================
|
160 |
-
# 4.
|
161 |
# ============================================================
|
162 |
-
|
163 |
def _perform_ocr(image_bytes):
|
164 |
if not OCR_ENABLED: return ""
|
165 |
try:
|
@@ -194,63 +205,47 @@ def _extract_from_docx(content_bytes):
|
|
194 |
def _extract_from_html(content_bytes, url):
|
195 |
soup = BeautifulSoup(content_bytes, "html.parser")
|
196 |
text_parts = []
|
197 |
-
# OCR على الصور داخل الصفحة
|
198 |
if OCR_ENABLED:
|
199 |
for img_tag in soup.find_all('img'):
|
200 |
img_src = img_tag.get('src')
|
201 |
if not img_src: continue
|
202 |
try:
|
203 |
img_url = urljoin(url, img_src)
|
204 |
-
img_resp =
|
205 |
-
img_resp
|
206 |
-
|
207 |
-
|
208 |
except Exception as e:
|
209 |
tqdm.write(f"⚠️ فشل تحميل أو معالجة الصورة {img_src}: {e}")
|
210 |
|
211 |
-
for element in soup(["script", "style", "nav", "footer", "header", "aside", "form", "button"]):
|
212 |
element.decompose()
|
213 |
-
|
214 |
-
|
|
|
215 |
|
216 |
def _clean_text(raw_text):
|
217 |
if not raw_text or not isinstance(raw_text, str):
|
218 |
return None
|
219 |
-
|
220 |
-
# إزالة محارف التحكم غير القابلة للطباعة
|
221 |
text = re.sub(r'[\x00-\x1f\x7f-\x9f]', '', raw_text)
|
222 |
-
# إزالة الروابط
|
223 |
text = re.sub(r'https?://\S+', '', text)
|
224 |
-
|
225 |
-
|
226 |
-
|
227 |
-
# تقسيم السطور وتنظيفها
|
228 |
-
clean_lines = [line.strip() for line in text.split('\n') if len(line.strip()) > 15]
|
229 |
-
if not clean_lines:
|
230 |
-
return None
|
231 |
-
|
232 |
-
final_text = "\n".join(clean_lines)
|
233 |
-
# تطبيع المسافات والأسطر الجديدة
|
234 |
final_text = re.sub(r" +", " ", final_text)
|
235 |
final_text = re.sub(r"(\n\s*){2,}", "\n\n", final_text)
|
236 |
-
|
237 |
return final_text.strip()
|
238 |
|
239 |
def extract_and_clean_text(url):
|
240 |
-
"""
|
241 |
-
الدالة الرئيسية الموجهة: تحمل المحتوى وتختار الطريقة المناسبة للاستخلاص.
|
242 |
-
"""
|
243 |
try:
|
244 |
-
|
245 |
-
resp
|
246 |
-
resp.raise_for_status()
|
247 |
content_bytes = resp.content
|
248 |
content_type = resp.headers.get('Content-Type', '').lower()
|
249 |
-
|
250 |
raw_text = ""
|
251 |
if url.lower().endswith(".pdf") or "application/pdf" in content_type:
|
252 |
raw_text = _extract_from_pdf(content_bytes)
|
253 |
-
elif url.lower().endswith(".docx") or "
|
254 |
raw_text = _extract_from_docx(content_bytes)
|
255 |
elif url.lower().endswith(".txt") or "text/plain" in content_type:
|
256 |
raw_text = content_bytes.decode('utf-8', errors='ignore')
|
@@ -258,20 +253,22 @@ def extract_and_clean_text(url):
|
|
258 |
raw_text = _extract_from_html(content_bytes, url)
|
259 |
else:
|
260 |
tqdm.write(f"⚠️ نوع ملف غير مدعوم في {url} ({content_type})")
|
261 |
-
return None
|
262 |
-
|
263 |
-
|
264 |
-
|
|
|
265 |
except Exception as e:
|
266 |
tqdm.write(f"❌ فشل استخراج وتنظيف النص من {url}: {e}")
|
267 |
-
return None
|
268 |
|
269 |
# ============================================================
|
270 |
-
# 5.
|
271 |
# ============================================================
|
272 |
-
def split_into_chunks(text, chunk_size=
|
273 |
if not text: return []
|
274 |
-
|
|
|
275 |
|
276 |
def get_embedding(text):
|
277 |
if GOOGLE_API_KEY:
|
@@ -282,208 +279,182 @@ def get_embedding(text):
|
|
282 |
return LOCAL_EMBEDDER.encode([text])[0].tolist()
|
283 |
|
284 |
# ============================================================
|
285 |
-
# 6.
|
286 |
# ============================================================
|
287 |
-
def
|
288 |
-
|
289 |
-
|
290 |
-
|
291 |
-
|
292 |
-
|
293 |
-
|
294 |
-
|
295 |
-
|
296 |
-
|
297 |
-
|
298 |
-
|
299 |
-
|
300 |
-
|
301 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
302 |
|
303 |
-
clean_content = extract_and_clean_text(link)
|
304 |
-
if not clean_content or len(clean_content) < 150:
|
305 |
-
tqdm.write(f"⚠️ محتوى غير كافٍ أو غير نظيف في {link}، سيتم تخطيه.")
|
306 |
-
continue
|
307 |
-
|
308 |
-
try:
|
309 |
-
summary = clean_content[:500].rsplit(' ', 1)[0] + '...'
|
310 |
cursor.execute(
|
311 |
-
"
|
312 |
-
(
|
|
|
|
|
|
|
313 |
)
|
314 |
document_id = cursor.lastrowid
|
315 |
-
|
316 |
chunks = split_into_chunks(clean_content)
|
317 |
for chunk in chunks:
|
318 |
embedding = get_embedding(chunk)
|
319 |
cursor.execute(
|
320 |
-
"INSERT INTO document_chunks (document_id, content, embedding) VALUES (%s, %s, %s)",
|
321 |
-
(document_id, chunk, json.dumps(embedding))
|
322 |
)
|
323 |
|
324 |
-
cursor.execute("UPDATE documents SET status = 'completed' WHERE id = %s", (document_id,))
|
325 |
conn.commit()
|
326 |
-
|
327 |
-
|
328 |
-
|
329 |
-
|
330 |
-
|
331 |
-
|
332 |
-
conn.rollback()
|
333 |
-
except Exception as e:
|
334 |
-
tqdm.write(f"❌ خطأ غير متوقع أثناء معالجة {link}: {e}")
|
335 |
-
conn.rollback()
|
336 |
|
337 |
-
|
338 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
339 |
return f"✅ اكتملت عملية المعالجة. تم حفظ {saved_docs_count} مستند جديد."
|
340 |
|
341 |
# ============================================================
|
342 |
-
# 7.
|
343 |
# ============================================================
|
|
|
344 |
is_running = False
|
345 |
|
346 |
def run_full_process_wrapper(max_documents_to_save, progress=gr.Progress(track_tqdm=True)):
|
347 |
-
global is_running
|
348 |
if is_running:
|
349 |
-
|
350 |
-
|
351 |
is_running = True
|
|
|
352 |
output_log = []
|
353 |
-
|
354 |
def log_message(msg):
|
355 |
nonlocal output_log
|
356 |
output_log.append(msg)
|
357 |
return "\n".join(output_log)
|
358 |
-
|
359 |
try:
|
360 |
-
yield log_message("🏁 بدء عملية الإعداد والتحقق...")
|
361 |
-
|
362 |
db_ok, db_msg = setup_database()
|
363 |
if not db_ok:
|
364 |
is_running = False
|
365 |
-
yield log_message(f"❌ فشل إعداد قاعدة البيانات: {db_msg}")
|
366 |
return
|
367 |
-
|
368 |
-
if not GOOGLE_API_KEY:
|
369 |
-
yield log_message("⚠️ تحذير: مفتاح Google API غير موجود. سيتم الاعتماد على النموذج المحلي.")
|
370 |
-
|
371 |
-
if not OCR_ENABLED:
|
372 |
-
yield log_message("⚠️ تحذير: ميزة OCR معطلة. لن يتم قراءة النصوص من الصور.")
|
373 |
-
|
374 |
-
if not DOCX_ENABLED:
|
375 |
-
yield log_message("⚠️ تحذير: ميزة قراءة DOCX معطلة.")
|
376 |
-
|
377 |
-
yield log_message(f"🚀 بدء عملية الزحف للموقع (حد وقائي {PAGE_LIMIT} صفحة)...")
|
378 |
-
|
379 |
all_links = crawl_site(SCRAPE_URL, page_limit=PAGE_LIMIT)
|
380 |
-
|
381 |
-
yield log_message(f"✅ تم العثور على {len(all_links)} رابط فريد.")
|
382 |
-
|
383 |
if not all_links:
|
384 |
is_running = False
|
385 |
-
yield log_message("⚠️ لم يتم العثور على أي روابط. انتهت العملية.")
|
386 |
return
|
387 |
-
|
388 |
max_docs_int = int(max_documents_to_save)
|
389 |
-
yield log_message(f"🔄 ستبدأ الآن عملية المعالجة والحفظ (الهدف: {max_docs_int} مستند)
|
390 |
-
|
391 |
-
|
392 |
-
yield log_message(f"\n🎉 {final_message}")
|
393 |
-
|
394 |
except Exception as e:
|
395 |
-
yield log_message(f"\n💥 حدث خطأ فادح وغير متوقع: {e}")
|
396 |
finally:
|
397 |
is_running = False
|
398 |
-
|
399 |
-
|
|
|
|
|
|
|
|
|
|
|
400 |
|
401 |
-
# --- NEW --- Function to list project files
|
402 |
def list_project_files():
|
403 |
-
"""Lists files in the current directory, returning a Pandas DataFrame."""
|
404 |
try:
|
405 |
files = os.listdir('.')
|
406 |
file_details = []
|
407 |
-
for f in files:
|
408 |
if os.path.isfile(f):
|
409 |
size_bytes = os.path.getsize(f)
|
410 |
size_kb = f"{size_bytes / 1024:.2f} KB"
|
411 |
file_details.append([f, size_kb])
|
412 |
-
|
413 |
if not file_details:
|
414 |
return pd.DataFrame(columns=["File Name", "Size"])
|
415 |
-
|
416 |
df = pd.DataFrame(file_details, columns=["اسم الملف", "الحجم"])
|
417 |
return df
|
418 |
except Exception as e:
|
419 |
return pd.DataFrame([["خطأ في عرض الملفات", str(e)]], columns=["اسم الملف", "الحجم"])
|
420 |
|
421 |
-
|
422 |
def main():
|
423 |
with gr.Blocks(theme=gr.themes.Soft(), title="Web Scraper & Indexer") as demo:
|
424 |
gr.Markdown(
|
425 |
"""
|
426 |
-
# 🚀 Web Scraper & Intelligent Indexer (
|
427 |
أداة متكاملة للتحكم في عملية كشط المواقع وفهرسة محتواها في قاعدة البيانات.
|
428 |
"""
|
429 |
)
|
430 |
-
|
431 |
-
# --- NEW --- Tabbed interface
|
432 |
with gr.Tabs():
|
433 |
-
# --- Main App Tab ---
|
434 |
with gr.Tab("التطبيق الرئيسي"):
|
435 |
-
with gr.
|
436 |
-
|
437 |
-
|
438 |
-
|
439 |
-
|
440 |
-
max_docs_input = gr.Number(
|
441 |
-
label="الحد الأقصى لعدد المستندات الجديدة للحفظ",
|
442 |
-
value=50,
|
443 |
-
minimum=1,
|
444 |
-
step=10
|
445 |
-
)
|
446 |
-
|
447 |
-
# --- NEW --- Collapsible log
|
448 |
-
with gr.Accordion("عرض سجل العمليات (Log)", open=True):
|
449 |
-
output_log = gr.Textbox(
|
450 |
-
label="سجل العمليات",
|
451 |
-
lines=20,
|
452 |
-
interactive=False,
|
453 |
-
autoscroll=True
|
454 |
-
)
|
455 |
|
456 |
-
|
457 |
-
|
458 |
-
|
459 |
-
|
460 |
-
|
461 |
-
|
462 |
-
|
463 |
-
|
464 |
-
|
465 |
-
|
466 |
-
|
467 |
-
)
|
468 |
|
469 |
-
# --- Files Tab ---
|
470 |
with gr.Tab("ملفات المشروع"):
|
471 |
-
|
472 |
-
|
473 |
-
|
474 |
-
|
475 |
-
|
476 |
-
|
477 |
-
|
478 |
-
|
479 |
-
fn=list_project_files,
|
480 |
-
inputs=[],
|
481 |
-
outputs=file_display
|
482 |
-
)
|
483 |
-
|
484 |
-
|
485 |
print("\n✅ الواجهة المحسنة جاهزة.")
|
486 |
-
demo.launch()
|
487 |
-
|
488 |
if __name__ == "__main__":
|
489 |
main()
|
|
|
2 |
import os
|
3 |
import re
|
4 |
import json
|
|
|
5 |
import io
|
6 |
+
import time
|
7 |
+
import concurrent.futures
|
8 |
import fitz # PyMuPDF
|
9 |
import requests
|
10 |
import google.generativeai as genai
|
11 |
import mysql.connector
|
12 |
+
from mysql.connector import pooling
|
13 |
from bs4 import BeautifulSoup
|
14 |
from dotenv import load_dotenv
|
15 |
from tqdm import tqdm
|
16 |
from sentence_transformers import SentenceTransformer
|
17 |
import gradio as gr
|
18 |
+
from urllib.parse import urljoin, urlparse, unquote
|
19 |
from collections import deque
|
20 |
+
import pandas as pd
|
21 |
|
22 |
+
# --- OCR and DOCX Libraries ---
|
23 |
try:
|
24 |
import pytesseract
|
25 |
from PIL import Image
|
|
|
37 |
DOCX_ENABLED = False
|
38 |
print("⚠️ تحذير: مكتبة python-docx غير موجودة. لن يتم قراءة ملفات وورد.")
|
39 |
|
|
|
40 |
# ============================================================
|
41 |
+
# 1. Settings and Global Variables
|
42 |
# ============================================================
|
43 |
print("⏳ جاري تحميل الإعدادات...")
|
44 |
if os.path.exists('.env'):
|
|
|
49 |
"user": os.getenv("DB_USER"),
|
50 |
"password": os.getenv("DB_PASSWORD"),
|
51 |
"database": os.getenv("DB_NAME"),
|
52 |
+
"charset": "utf8mb4"
|
53 |
}
|
54 |
|
55 |
+
try:
|
56 |
+
print("⏳ جاري إنشاء مجمع الاتصالات بقاعدة البيانات...")
|
57 |
+
connection_pool = mysql.connector.pooling.MySQLConnectionPool(
|
58 |
+
pool_name="scraper_pool",
|
59 |
+
pool_size=10,
|
60 |
+
**DB_CONFIG
|
61 |
+
)
|
62 |
+
print("✅ تم إنشاء مجمع الاتصالات بنجاح.")
|
63 |
+
except mysql.connector.Error as err:
|
64 |
+
print(f"❌ فشل فادح في إنشاء مجمع الاتصالات: {err}")
|
65 |
+
exit()
|
66 |
+
|
67 |
GOOGLE_API_KEY = os.getenv("GOOGLE_API_KEY")
|
68 |
if GOOGLE_API_KEY:
|
69 |
try:
|
|
|
77 |
|
78 |
SCRAPE_URL = os.getenv("SCRAPE_URL", "https://fra.gov.eg/")
|
79 |
REQUEST_TIMEOUT = 60
|
80 |
+
PAGE_LIMIT = 500
|
81 |
+
MAX_WORKERS = 5
|
82 |
|
83 |
+
print(f"⏳ جاري تحميل نموذج Embedding المحلي... (قد يستغرق بعض الوقت أول مرة)")
|
84 |
LOCAL_EMBEDDER = SentenceTransformer("all-MiniLM-L6-v2")
|
85 |
print("✅ تم تحميل نموذج Embedding المحلي.")
|
86 |
|
87 |
# ============================================================
|
88 |
+
# 2. Database and Network Helpers
|
89 |
# ============================================================
|
90 |
def setup_database():
|
91 |
+
"""Checks if the database connection from the pool is valid."""
|
92 |
try:
|
93 |
+
with connection_pool.get_connection() as conn:
|
94 |
+
with conn.cursor(buffered=True) as cursor:
|
95 |
+
cursor.execute("SELECT 1")
|
96 |
+
cursor.fetchone()
|
97 |
+
print("✅ الاتصال من المجمع (Pool) سليم وجاهز.")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
98 |
return True, "قاعدة البيانات جاهزة."
|
99 |
except mysql.connector.Error as err:
|
100 |
+
error_msg = f"❌ خطأ فادح في قاعدة البيانات: {err}"
|
101 |
+
print(error_msg)
|
102 |
+
return False, str(err)
|
103 |
+
|
104 |
+
def get_document_count():
|
105 |
+
"""Connects to the DB via the pool and returns the current number of documents."""
|
106 |
+
try:
|
107 |
+
with connection_pool.get_connection() as conn:
|
108 |
+
with conn.cursor(buffered=True) as cursor:
|
109 |
+
cursor.execute("SELECT COUNT(*) FROM documents")
|
110 |
+
count = cursor.fetchone()[0]
|
111 |
+
return f"### **العدد الحالي:** {count} مستند"
|
112 |
+
except Exception as e:
|
113 |
+
print(f"Error getting document count: {e}")
|
114 |
+
return "### **العدد الحالي:** خطأ في الاتصال بقاعدة البيانات"
|
115 |
+
|
116 |
+
def requests_with_retry(url, method='get', retries=3, delay=5, **kwargs):
|
117 |
+
headers = kwargs.get("headers", {"User-Agent": "Mozilla/5.0"})
|
118 |
+
kwargs["headers"] = headers
|
119 |
+
for i in range(retries):
|
120 |
+
try:
|
121 |
+
if method.lower() == 'get':
|
122 |
+
response = requests.get(url, **kwargs)
|
123 |
+
elif method.lower() == 'head':
|
124 |
+
response = requests.head(url, **kwargs)
|
125 |
+
else:
|
126 |
+
raise ValueError("Unsupported HTTP method")
|
127 |
+
response.raise_for_status()
|
128 |
+
return response
|
129 |
+
except requests.exceptions.RequestException as e:
|
130 |
+
tqdm.write(f"⚠️ فشلت محاولة {i+1} للرابط {url}. خطأ: {e}. سيتم المحاولة مرة أخرى بعد {delay} ثانية.")
|
131 |
+
if i < retries - 1:
|
132 |
+
time.sleep(delay)
|
133 |
+
tqdm.write(f"❌ فشلت كل المحاولات ({retries}) للرابط {url}.")
|
134 |
+
return None
|
135 |
|
136 |
# ============================================================
|
137 |
+
# 3. Crawler Functions
|
138 |
# ============================================================
|
139 |
+
def crawl_site(start_url, page_limit=100):
|
140 |
visited = set([start_url])
|
141 |
queue = deque([start_url])
|
142 |
all_links = set([start_url])
|
143 |
base_domain = urlparse(start_url).netloc
|
|
|
144 |
with tqdm(total=page_limit, desc="🔍 Crawling Progress", unit="page") as pbar:
|
145 |
while queue and len(visited) < page_limit:
|
146 |
url = queue.popleft()
|
147 |
pbar.set_description(f"🔍 Crawling: {url[:70]}...")
|
|
|
148 |
try:
|
149 |
+
resp_head = requests_with_retry(url, method='head', timeout=15, allow_redirects=True)
|
150 |
+
if not resp_head: continue
|
151 |
+
content_type = resp_head.headers.get("Content-Type", "").lower()
|
|
|
152 |
if "text/html" in content_type:
|
153 |
+
resp_get = requests_with_retry(url, timeout=REQUEST_TIMEOUT)
|
154 |
+
if not resp_get: continue
|
155 |
+
soup = BeautifulSoup(resp_get.text, "html.parser")
|
156 |
for a_tag in soup.find_all("a", href=True):
|
157 |
href = urljoin(url, a_tag["href"]).split("#")[0].strip()
|
158 |
if not href.startswith("http") or urlparse(href).netloc != base_domain or href in visited:
|
159 |
continue
|
160 |
if len(visited) >= page_limit: break
|
|
|
161 |
visited.add(href)
|
162 |
all_links.add(href)
|
163 |
pbar.update(1)
|
|
|
164 |
if not any(href.lower().endswith(ext) for ext in ['.pdf', '.docx', '.txt']):
|
165 |
queue.append(href)
|
166 |
+
except Exception as e:
|
167 |
+
tqdm.write(f"⚠️ خطأ غير متوقع أثناء الزحف إلى {url}: {e}")
|
168 |
continue
|
169 |
return list(all_links)
|
170 |
|
171 |
# ============================================================
|
172 |
+
# 4. Text Extraction Functions
|
173 |
# ============================================================
|
|
|
174 |
def _perform_ocr(image_bytes):
|
175 |
if not OCR_ENABLED: return ""
|
176 |
try:
|
|
|
205 |
def _extract_from_html(content_bytes, url):
|
206 |
soup = BeautifulSoup(content_bytes, "html.parser")
|
207 |
text_parts = []
|
|
|
208 |
if OCR_ENABLED:
|
209 |
for img_tag in soup.find_all('img'):
|
210 |
img_src = img_tag.get('src')
|
211 |
if not img_src: continue
|
212 |
try:
|
213 |
img_url = urljoin(url, img_src)
|
214 |
+
img_resp = requests_with_retry(img_url, timeout=15)
|
215 |
+
if img_resp:
|
216 |
+
tqdm.write(f"📝 OCR on image from URL: {img_url[:80]}...")
|
217 |
+
text_parts.append(_perform_ocr(img_resp.content))
|
218 |
except Exception as e:
|
219 |
tqdm.write(f"⚠️ فشل تحميل أو معالجة الصورة {img_src}: {e}")
|
220 |
|
221 |
+
for element in soup(["script", "style", "nav", "footer", "header", "aside", "form", "button", "noscript"]):
|
222 |
element.decompose()
|
223 |
+
main_content = soup.find("main") or soup.find("article") or soup.find("body")
|
224 |
+
text = main_content.get_text(separator='\n', strip=True) if main_content else ""
|
225 |
+
return text + "\n" + "\n".join(text_parts)
|
226 |
|
227 |
def _clean_text(raw_text):
|
228 |
if not raw_text or not isinstance(raw_text, str):
|
229 |
return None
|
|
|
|
|
230 |
text = re.sub(r'[\x00-\x1f\x7f-\x9f]', '', raw_text)
|
|
|
231 |
text = re.sub(r'https?://\S+', '', text)
|
232 |
+
lines = [line.strip() for line in text.split('\n') if len(line.strip()) > 10]
|
233 |
+
if not lines: return None
|
234 |
+
final_text = "\n".join(lines)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
235 |
final_text = re.sub(r" +", " ", final_text)
|
236 |
final_text = re.sub(r"(\n\s*){2,}", "\n\n", final_text)
|
|
|
237 |
return final_text.strip()
|
238 |
|
239 |
def extract_and_clean_text(url):
|
|
|
|
|
|
|
240 |
try:
|
241 |
+
resp = requests_with_retry(url, timeout=REQUEST_TIMEOUT)
|
242 |
+
if not resp: return None, None
|
|
|
243 |
content_bytes = resp.content
|
244 |
content_type = resp.headers.get('Content-Type', '').lower()
|
|
|
245 |
raw_text = ""
|
246 |
if url.lower().endswith(".pdf") or "application/pdf" in content_type:
|
247 |
raw_text = _extract_from_pdf(content_bytes)
|
248 |
+
elif url.lower().endswith(".docx") or "vnd.openxmlformats-officedocument.wordprocessingml.document" in content_type:
|
249 |
raw_text = _extract_from_docx(content_bytes)
|
250 |
elif url.lower().endswith(".txt") or "text/plain" in content_type:
|
251 |
raw_text = content_bytes.decode('utf-8', errors='ignore')
|
|
|
253 |
raw_text = _extract_from_html(content_bytes, url)
|
254 |
else:
|
255 |
tqdm.write(f"⚠️ نوع ملف غير مدعوم في {url} ({content_type})")
|
256 |
+
return None, None
|
257 |
+
cleaned = _clean_text(raw_text)
|
258 |
+
if cleaned and len(cleaned) >= 50:
|
259 |
+
return cleaned, url
|
260 |
+
return None, None
|
261 |
except Exception as e:
|
262 |
tqdm.write(f"❌ فشل استخراج وتنظيف النص من {url}: {e}")
|
263 |
+
return None, None
|
264 |
|
265 |
# ============================================================
|
266 |
+
# 5. Embeddings
|
267 |
# ============================================================
|
268 |
+
def split_into_chunks(text, chunk_size=500, overlap=100):
|
269 |
if not text: return []
|
270 |
+
step = max(1, chunk_size - overlap)
|
271 |
+
return [text[i:i + chunk_size] for i in range(0, len(text), step)]
|
272 |
|
273 |
def get_embedding(text):
|
274 |
if GOOGLE_API_KEY:
|
|
|
279 |
return LOCAL_EMBEDDER.encode([text])[0].tolist()
|
280 |
|
281 |
# ============================================================
|
282 |
+
# 6. Database Storage
|
283 |
# ============================================================
|
284 |
+
def process_single_link(link):
|
285 |
+
"""Processes a single link using a connection from the pool."""
|
286 |
+
try:
|
287 |
+
with connection_pool.get_connection() as conn:
|
288 |
+
with conn.cursor(buffered=True) as cursor:
|
289 |
+
cursor.execute("SELECT id FROM documents WHERE original_path = %s", (link,))
|
290 |
+
if cursor.fetchone():
|
291 |
+
return f"🔄 الرابط {link} موجود بالفعل، سيتم تخطيه."
|
292 |
+
|
293 |
+
clean_content, source_url = extract_and_clean_text(link)
|
294 |
+
if not clean_content: return None
|
295 |
+
|
296 |
+
summary = clean_content[:450] + '...' if len(clean_content) > 450 else clean_content
|
297 |
+
|
298 |
+
# --- FINAL FIX 1: Decode filename to handle Arabic in URLs ---
|
299 |
+
file_name_encoded = os.path.basename(urlparse(link).path)
|
300 |
+
file_name = unquote(file_name_encoded)
|
301 |
+
if not file_name: # If the path is empty (like for the base domain)
|
302 |
+
file_name = link # Use the full link as a fallback
|
303 |
+
# ---
|
304 |
+
|
305 |
+
# --- FINAL FIX 2: Change status to 'completed' to match Laravel ENUM ---
|
306 |
+
status_to_insert = 'completed'
|
307 |
+
# ---
|
308 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
309 |
cursor.execute(
|
310 |
+
"""
|
311 |
+
INSERT INTO documents (filename, original_path, summary, status, created_at, updated_at)
|
312 |
+
VALUES (%s, %s, %s, %s, NOW(), NOW())
|
313 |
+
""",
|
314 |
+
(file_name, link, summary, status_to_insert)
|
315 |
)
|
316 |
document_id = cursor.lastrowid
|
317 |
+
|
318 |
chunks = split_into_chunks(clean_content)
|
319 |
for chunk in chunks:
|
320 |
embedding = get_embedding(chunk)
|
321 |
cursor.execute(
|
322 |
+
"INSERT INTO document_chunks (document_id, content, embedding, created_at, updated_at) VALUES (%s, %s, %s, NOW(), NOW())",
|
323 |
+
(document_id, chunk, json.dumps(embedding) if embedding else None)
|
324 |
)
|
325 |
|
|
|
326 |
conn.commit()
|
327 |
+
# Use the clean file_name in the success message for clarity
|
328 |
+
return f"✅ تم حفظ '{file_name}' بنجاح مع {len(chunks)} جزء."
|
329 |
+
except mysql.connector.Error as err:
|
330 |
+
return f"❌ خطأ قاعدة بيانات أثناء معالجة {link}: {err}"
|
331 |
+
except Exception as e:
|
332 |
+
return f"❌ خطأ غير متوقع أثناء معالجة {link}: {e}"
|
|
|
|
|
|
|
|
|
333 |
|
334 |
+
def process_and_store_links(links, max_docs, stop_flag):
|
335 |
+
saved_docs_count = 0
|
336 |
+
with tqdm(total=max_docs, desc="💾 Saving Documents") as pbar:
|
337 |
+
with concurrent.futures.ThreadPoolExecutor(max_workers=MAX_WORKERS) as executor:
|
338 |
+
future_to_link = {executor.submit(process_single_link, link): link for link in links}
|
339 |
+
for future in concurrent.futures.as_completed(future_to_link):
|
340 |
+
if stop_flag["stop"] or saved_docs_count >= max_docs:
|
341 |
+
for f in future_to_link: f.cancel()
|
342 |
+
break
|
343 |
+
try:
|
344 |
+
result_message = future.result()
|
345 |
+
if result_message:
|
346 |
+
tqdm.write(result_message)
|
347 |
+
if result_message.startswith("✅"):
|
348 |
+
saved_docs_count += 1
|
349 |
+
pbar.update(1)
|
350 |
+
except Exception as exc:
|
351 |
+
link = future_to_link[future]
|
352 |
+
tqdm.write(f"💥 حدث خطأ فادح في معالج الرابط {link}: {exc}")
|
353 |
return f"✅ اكتملت عملية المعالجة. تم حفظ {saved_docs_count} مستند جديد."
|
354 |
|
355 |
# ============================================================
|
356 |
+
# 7. Gradio UI and Main Execution
|
357 |
# ============================================================
|
358 |
+
stop_flag = {"stop": False}
|
359 |
is_running = False
|
360 |
|
361 |
def run_full_process_wrapper(max_documents_to_save, progress=gr.Progress(track_tqdm=True)):
|
362 |
+
global is_running, stop_flag
|
363 |
if is_running:
|
364 |
+
yield "⏳ العملية قيد التشغيل بالفعل...", "⏳ قيد التشغيل..."
|
365 |
+
return
|
366 |
is_running = True
|
367 |
+
stop_flag["stop"] = False
|
368 |
output_log = []
|
|
|
369 |
def log_message(msg):
|
370 |
nonlocal output_log
|
371 |
output_log.append(msg)
|
372 |
return "\n".join(output_log)
|
|
|
373 |
try:
|
374 |
+
yield log_message("🏁 بدء عملية الإعداد والتحقق..."), "⏳ قيد التشغيل..."
|
|
|
375 |
db_ok, db_msg = setup_database()
|
376 |
if not db_ok:
|
377 |
is_running = False
|
378 |
+
yield log_message(f"❌ فشل إعداد قاعدة البيانات: {db_msg}"), "🔴 فشل"
|
379 |
return
|
380 |
+
yield log_message(f"🚀 بدء عملية الزحف للموقع (حد وقائي {PAGE_LIMIT} صفحة)..."), "⏳ قيد التشغيل..."
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
381 |
all_links = crawl_site(SCRAPE_URL, page_limit=PAGE_LIMIT)
|
382 |
+
yield log_message(f"✅ تم العثور على {len(all_links)} رابط فريد."), "⏳ قيد التشغيل..."
|
|
|
|
|
383 |
if not all_links:
|
384 |
is_running = False
|
385 |
+
yield log_message("⚠️ لم يتم العثور على أي روابط. انتهت العملية."), "✅ مكتمل"
|
386 |
return
|
|
|
387 |
max_docs_int = int(max_documents_to_save)
|
388 |
+
yield log_message(f"🔄 ستبدأ الآن عملية المعالجة والحفظ (الهدف: {max_docs_int} مستند) باستخدام {MAX_WORKERS} معالج متوازي."), "⏳ قيد التشغيل..."
|
389 |
+
final_message = process_and_store_links(all_links, max_docs_int, stop_flag)
|
390 |
+
yield log_message(f"\n🎉 {final_message}"), "✅ مكتمل"
|
|
|
|
|
391 |
except Exception as e:
|
392 |
+
yield log_message(f"\n💥 حدث خطأ فادح وغير متوقع: {e}"), "🔴 فشل"
|
393 |
finally:
|
394 |
is_running = False
|
395 |
+
if stop_flag["stop"]:
|
396 |
+
yield log_message("⏹️ تم إيقاف العملية."), "🛑 متوقف"
|
397 |
+
|
398 |
+
def stop_process():
|
399 |
+
global stop_flag
|
400 |
+
stop_flag["stop"] = True
|
401 |
+
return "🛑 جاري إيقاف العملية...", "🛑 متوقف"
|
402 |
|
|
|
403 |
def list_project_files():
|
|
|
404 |
try:
|
405 |
files = os.listdir('.')
|
406 |
file_details = []
|
407 |
+
for f in sorted(files):
|
408 |
if os.path.isfile(f):
|
409 |
size_bytes = os.path.getsize(f)
|
410 |
size_kb = f"{size_bytes / 1024:.2f} KB"
|
411 |
file_details.append([f, size_kb])
|
|
|
412 |
if not file_details:
|
413 |
return pd.DataFrame(columns=["File Name", "Size"])
|
|
|
414 |
df = pd.DataFrame(file_details, columns=["اسم الملف", "الحجم"])
|
415 |
return df
|
416 |
except Exception as e:
|
417 |
return pd.DataFrame([["خطأ في عرض الملفات", str(e)]], columns=["اسم الملف", "الحجم"])
|
418 |
|
|
|
419 |
def main():
|
420 |
with gr.Blocks(theme=gr.themes.Soft(), title="Web Scraper & Indexer") as demo:
|
421 |
gr.Markdown(
|
422 |
"""
|
423 |
+
# 🚀 Web Scraper & Intelligent Indexer (V9 - Final)
|
424 |
أداة متكاملة للتحكم في عملية كشط المواقع وفهرسة محتواها في قاعدة البيانات.
|
425 |
"""
|
426 |
)
|
|
|
|
|
427 |
with gr.Tabs():
|
|
|
428 |
with gr.Tab("التطبيق الرئيسي"):
|
429 |
+
with gr.Row():
|
430 |
+
status_indicator = gr.Textbox(label="الحالة", value="⚪ خامل", interactive=False, scale=3)
|
431 |
+
doc_count_display = gr.Markdown("### العدد الحالي: جاري التحميل...")
|
432 |
+
refresh_count_btn = gr.Button("🔄 تحديث العدد")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
433 |
|
434 |
+
with gr.Row():
|
435 |
+
run_btn = gr.Button("🚀 ابدأ الزحف والفهرسة", variant="primary", scale=2)
|
436 |
+
stop_btn = gr.Button("🛑 إيقاف", variant="stop", scale=1)
|
437 |
+
|
438 |
+
max_docs_input = gr.Number(label="الحد الأقصى لعدد المستندات الجديدة للإضافة", value=50, minimum=1, step=10)
|
439 |
+
|
440 |
+
with gr.Accordion("عرض سجل العمليات (Log)", open=True):
|
441 |
+
output_log = gr.Textbox(label="سجل العمليات", lines=20, interactive=False, autoscroll=True)
|
442 |
+
|
443 |
+
run_event = run_btn.click(fn=run_full_process_wrapper, inputs=[max_docs_input], outputs=[output_log, status_indicator])
|
444 |
+
stop_btn.click(fn=stop_process, inputs=[], outputs=[output_log, status_indicator], cancels=[run_event])
|
445 |
+
refresh_count_btn.click(fn=get_document_count, inputs=None, outputs=doc_count_display)
|
446 |
|
|
|
447 |
with gr.Tab("ملفات المشروع"):
|
448 |
+
gr.Markdown("## عرض الملفات الموجودة في مجلد المشروع.")
|
449 |
+
refresh_files_btn = gr.Button("🔄 تحديث قائمة الملفات")
|
450 |
+
file_display = gr.DataFrame(headers=["اسم الملف", "الحجم"], datatype=["str", "str"], label="قائمة الملفات")
|
451 |
+
refresh_files_btn.click(fn=list_project_files, inputs=[], outputs=file_display)
|
452 |
+
|
453 |
+
demo.load(fn=get_document_count, inputs=None, outputs=doc_count_display)
|
454 |
+
demo.load(fn=list_project_files, inputs=[], outputs=file_display)
|
455 |
+
|
|
|
|
|
|
|
|
|
|
|
|
|
456 |
print("\n✅ الواجهة المحسنة جاهزة.")
|
457 |
+
demo.launch(server_name="0.0.0.0", share=True)
|
458 |
+
|
459 |
if __name__ == "__main__":
|
460 |
main()
|