Spaces:
Sleeping
Sleeping
Create app.py
Browse files
app.py
ADDED
|
@@ -0,0 +1,112 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import gradio as gr
|
| 2 |
+
import requests
|
| 3 |
+
from bs4 import BeautifulSoup
|
| 4 |
+
from googlesearch import search
|
| 5 |
+
from huggingface_hub import InferenceClient
|
| 6 |
+
from langdetect import detect
|
| 7 |
+
import os
|
| 8 |
+
|
| 9 |
+
# إعداد العميل (استبدل TOKEN بـ Secret في إعدادات HF إذا لزم الأمر، أو سيتم استخدامه تلقائياً إذا كان الموديل عاماً)
|
| 10 |
+
# سنستخدم Gemma 3 4B لسرعته وجودته العالية
|
| 11 |
+
client = InferenceClient("google/gemma-3-4b-it")
|
| 12 |
+
|
| 13 |
+
def get_language(text):
|
| 14 |
+
try:
|
| 15 |
+
return detect(text)
|
| 16 |
+
except:
|
| 17 |
+
return "ar"
|
| 18 |
+
|
| 19 |
+
def search_islamweb(query):
|
| 20 |
+
# حصر البحث في إسلام ويب فقط
|
| 21 |
+
full_query = f"site:islamweb.net {query}"
|
| 22 |
+
try:
|
| 23 |
+
links = [j for j in search(full_query, num=3, stop=3, pause=2)]
|
| 24 |
+
return links
|
| 25 |
+
except:
|
| 26 |
+
return []
|
| 27 |
+
|
| 28 |
+
def scrape_content(url):
|
| 29 |
+
try:
|
| 30 |
+
headers = {'User-Agent': 'Mozilla/5.0'}
|
| 31 |
+
response = requests.get(url, headers=headers, timeout=10)
|
| 32 |
+
response.encoding = 'utf-8'
|
| 33 |
+
soup = BeautifulSoup(response.text, 'html.parser')
|
| 34 |
+
|
| 35 |
+
# استخراج متن الفتوى (هذا الـ Selector مناسب لبنية إسلام ويب الحالية)
|
| 36 |
+
main_content = soup.find('div', {'class': 'item'})
|
| 37 |
+
if not main_content:
|
| 38 |
+
main_content = soup.find('div', {'id': 'fullitem'})
|
| 39 |
+
|
| 40 |
+
return main_content.get_text(strip=True)[:4000] # نأخذ أول 4000 حرف للسياق
|
| 41 |
+
except:
|
| 42 |
+
return ""
|
| 43 |
+
|
| 44 |
+
def islamic_ai_search(question):
|
| 45 |
+
lang = get_language(question)
|
| 46 |
+
|
| 47 |
+
# 1. البحث عن الروابط
|
| 48 |
+
links = search_islamweb(question)
|
| 49 |
+
if not links:
|
| 50 |
+
return "عذراً، لم أجد نتائج في إسلام ويب لهذا السؤال. / Üzgünüm, bu soru için IslamWeb'de sonuç bulamadım.", ""
|
| 51 |
+
|
| 52 |
+
# 2. استخراج السياق من أول رابط ناجح
|
| 53 |
+
context = ""
|
| 54 |
+
source_link = ""
|
| 55 |
+
for link in links:
|
| 56 |
+
content = scrape_content(link)
|
| 57 |
+
if len(content) > 100:
|
| 58 |
+
context = content
|
| 59 |
+
source_link = link
|
| 60 |
+
break
|
| 61 |
+
|
| 62 |
+
if not context:
|
| 63 |
+
return "تعذر استخراج محتوى مناسب من المصدر.", ""
|
| 64 |
+
|
| 65 |
+
# 3. صياغة البرومبت المحكم (System Prompt) لعدم الهلوسة
|
| 66 |
+
system_message = (
|
| 67 |
+
"You are an expert Islamic researcher. Answer the user's question ONLY using the provided context from IslamWeb. "
|
| 68 |
+
"Strict rules: \n"
|
| 69 |
+
"1. If the answer is not in the context, say you don't know.\n"
|
| 70 |
+
"2. You MUST answer in the same language as the user's question.\n"
|
| 71 |
+
"3. Do not add outside information.\n"
|
| 72 |
+
f"Context: {context}"
|
| 73 |
+
)
|
| 74 |
+
|
| 75 |
+
# 4. طلب الإجابة من النموذج
|
| 76 |
+
response = ""
|
| 77 |
+
try:
|
| 78 |
+
for message in client.chat_completion(
|
| 79 |
+
messages=[
|
| 80 |
+
{"role": "system", "content": system_message},
|
| 81 |
+
{"role": "user", "content": question}
|
| 82 |
+
],
|
| 83 |
+
max_tokens=1000,
|
| 84 |
+
stream=True,
|
| 85 |
+
):
|
| 86 |
+
token = message.choices[0].delta.content
|
| 87 |
+
response += token
|
| 88 |
+
yield response, source_link
|
| 89 |
+
except Exception as e:
|
| 90 |
+
yield f"خطأ في الاتصال بالنموذج: {str(e)}", source_link
|
| 91 |
+
|
| 92 |
+
# إعداد واجهة Gradio
|
| 93 |
+
with gr.Blocks(theme=gr.themes.Soft(primary_hue="green")) as demo:
|
| 94 |
+
gr.Markdown("# 🕋 الباحث الإسلامي الذكي (مرجع إسلام ويب)")
|
| 95 |
+
gr.Markdown("اسأل أي سؤال ديني، وسيقوم الذكاء الاصطناعي بالبحث في إسلام ويب وتلخيص الإجابة لك.")
|
| 96 |
+
|
| 97 |
+
with gr.Row():
|
| 98 |
+
input_text = gr.Textbox(label="سؤالك / Sorunuz", placeholder="مثلاً: ما حكم صلاة الوتر؟")
|
| 99 |
+
|
| 100 |
+
submit_btn = gr.Button("بحث وتحليل", variant="primary")
|
| 101 |
+
|
| 102 |
+
with gr.Column():
|
| 103 |
+
output_answer = gr.Markdown(label="الإجابة")
|
| 104 |
+
output_source = gr.Textbox(label="المصدر الأصلي (إسلام ويب)")
|
| 105 |
+
|
| 106 |
+
submit_btn.click(
|
| 107 |
+
fn=islamic_ai_search,
|
| 108 |
+
inputs=input_text,
|
| 109 |
+
outputs=[output_answer, output_source]
|
| 110 |
+
)
|
| 111 |
+
|
| 112 |
+
demo.queue().launch()
|