Spaces:
Runtime error
Runtime error
Upload 2 files
Browse files- app.py +45 -43
- requirements.txt +2 -2
app.py
CHANGED
|
@@ -7,9 +7,8 @@ from bidi.algorithm import get_display
|
|
| 7 |
import re
|
| 8 |
from collections import Counter
|
| 9 |
from transformers import AutoTokenizer, AutoModelForCausalLM
|
| 10 |
-
import
|
| 11 |
-
|
| 12 |
-
from duckduckgo_search import ddg
|
| 13 |
|
| 14 |
# تهيئة النموذج
|
| 15 |
try:
|
|
@@ -34,52 +33,55 @@ class ArticleGenerator:
|
|
| 34 |
"""البحث عن معلومات حول الموضوع"""
|
| 35 |
results = []
|
| 36 |
|
| 37 |
-
# البحث في ويكيبيديا
|
| 38 |
try:
|
| 39 |
-
|
| 40 |
-
|
| 41 |
-
|
| 42 |
-
|
| 43 |
-
|
| 44 |
-
|
| 45 |
-
|
| 46 |
-
|
| 47 |
-
|
| 48 |
-
|
| 49 |
-
|
| 50 |
-
|
| 51 |
-
|
| 52 |
-
|
| 53 |
-
|
| 54 |
-
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
| 59 |
-
|
| 60 |
-
'content': result['body']
|
| 61 |
-
})
|
| 62 |
except:
|
| 63 |
-
|
| 64 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 65 |
return results
|
| 66 |
|
| 67 |
def extract_keywords(self, topic):
|
| 68 |
"""استخراج الكلمات المفتاحية من الموضوع"""
|
| 69 |
-
|
| 70 |
-
|
| 71 |
-
|
| 72 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 73 |
|
| 74 |
-
|
| 75 |
-
outputs = self.model.generate(
|
| 76 |
-
inputs["input_ids"],
|
| 77 |
-
max_length=200,
|
| 78 |
-
temperature=0.7,
|
| 79 |
-
num_return_sequences=1
|
| 80 |
-
)
|
| 81 |
-
keywords = self.tokenizer.decode(outputs[0], skip_special_tokens=True)
|
| 82 |
-
return [kw.strip() for kw in keywords.split(',')]
|
| 83 |
|
| 84 |
def generate_content_with_research(self, topic, style):
|
| 85 |
"""توليد محتوى مبني على البحث"""
|
|
@@ -88,7 +90,7 @@ class ArticleGenerator:
|
|
| 88 |
|
| 89 |
# البحث عن كل كلمة مفتاحية
|
| 90 |
all_research = []
|
| 91 |
-
for keyword in keywords:
|
| 92 |
search_results = self.search_topic(f"{topic} {keyword}")
|
| 93 |
all_research.extend(search_results)
|
| 94 |
|
|
|
|
| 7 |
import re
|
| 8 |
from collections import Counter
|
| 9 |
from transformers import AutoTokenizer, AutoModelForCausalLM
|
| 10 |
+
import requests
|
| 11 |
+
from bs4 import BeautifulSoup
|
|
|
|
| 12 |
|
| 13 |
# تهيئة النموذج
|
| 14 |
try:
|
|
|
|
| 33 |
"""البحث عن معلومات حول الموضوع"""
|
| 34 |
results = []
|
| 35 |
|
|
|
|
| 36 |
try:
|
| 37 |
+
# استخدام محرك بحث عربي
|
| 38 |
+
search_url = f"https://www.google.com/search?q={topic}&hl=ar"
|
| 39 |
+
headers = {
|
| 40 |
+
'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36'
|
| 41 |
+
}
|
| 42 |
+
response = requests.get(search_url, headers=headers)
|
| 43 |
+
|
| 44 |
+
if response.status_code == 200:
|
| 45 |
+
soup = BeautifulSoup(response.text, 'html.parser')
|
| 46 |
+
search_results = soup.find_all('div', class_='g')
|
| 47 |
+
|
| 48 |
+
for result in search_results[:num_results]:
|
| 49 |
+
title_elem = result.find('h3')
|
| 50 |
+
snippet_elem = result.find('div', class_='VwiC3b')
|
| 51 |
+
|
| 52 |
+
if title_elem and snippet_elem:
|
| 53 |
+
results.append({
|
| 54 |
+
'source': 'web',
|
| 55 |
+
'title': title_elem.text,
|
| 56 |
+
'content': snippet_elem.text
|
| 57 |
+
})
|
|
|
|
|
|
|
| 58 |
except:
|
| 59 |
+
# في حالة فشل البحث، استخدم محتوى افتراضي
|
| 60 |
+
results.append({
|
| 61 |
+
'source': 'default',
|
| 62 |
+
'title': f'معلومات عن {topic}',
|
| 63 |
+
'content': f'يعتبر موضوع {topic} من المواضيع المهمة في وقتنا الحاضر.'
|
| 64 |
+
})
|
| 65 |
+
|
| 66 |
return results
|
| 67 |
|
| 68 |
def extract_keywords(self, topic):
|
| 69 |
"""استخراج الكلمات المفتاحية من الموضوع"""
|
| 70 |
+
# قائمة من الكلمات المفتاحية المحتملة
|
| 71 |
+
potential_keywords = [
|
| 72 |
+
f"تعريف {topic}",
|
| 73 |
+
f"أهمية {topic}",
|
| 74 |
+
f"فوائد {topic}",
|
| 75 |
+
f"أنواع {topic}",
|
| 76 |
+
f"مميزات {topic}",
|
| 77 |
+
f"تطبيقات {topic}",
|
| 78 |
+
f"تاريخ {topic}",
|
| 79 |
+
f"مستقبل {topic}",
|
| 80 |
+
f"تحديات {topic}",
|
| 81 |
+
f"حلول {topic}"
|
| 82 |
+
]
|
| 83 |
|
| 84 |
+
return potential_keywords
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 85 |
|
| 86 |
def generate_content_with_research(self, topic, style):
|
| 87 |
"""توليد محتوى مبني على البحث"""
|
|
|
|
| 90 |
|
| 91 |
# البحث عن كل كلمة مفتاحية
|
| 92 |
all_research = []
|
| 93 |
+
for keyword in keywords[:3]: # نأخذ أول 3 كلمات مفتاحية فقط
|
| 94 |
search_results = self.search_topic(f"{topic} {keyword}")
|
| 95 |
all_research.extend(search_results)
|
| 96 |
|
requirements.txt
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
transformers>=4.30.0
|
| 2 |
torch>=2.0.0
|
| 3 |
-
|
| 4 |
-
|
| 5 |
gradio>=4.0.0
|
| 6 |
python-dotenv==1.0.0
|
| 7 |
sentencepiece==0.1.99
|
|
|
|
| 1 |
transformers>=4.30.0
|
| 2 |
torch>=2.0.0
|
| 3 |
+
requests>=2.25.1
|
| 4 |
+
beautifulsoup4>=4.9.3
|
| 5 |
gradio>=4.0.0
|
| 6 |
python-dotenv==1.0.0
|
| 7 |
sentencepiece==0.1.99
|