Spaces:
Sleeping
Sleeping
Update helpers.py
Browse files- helpers.py +27 -1
helpers.py
CHANGED
|
@@ -15,7 +15,33 @@ import shutil
|
|
| 15 |
import requests
|
| 16 |
from bs4 import BeautifulSoup
|
| 17 |
|
| 18 |
-
from file_loader import get_vectorstore
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 19 |
|
| 20 |
async def get_urls_splits(url='https://nct.neu.edu.vn/', char='https://nct.neu.edu.vn/'):
|
| 21 |
reqs = requests.get(url)
|
|
|
|
| 15 |
import requests
|
| 16 |
from bs4 import BeautifulSoup
|
| 17 |
|
| 18 |
+
# from file_loader import get_vectorstore
|
| 19 |
+
|
| 20 |
+
###
|
| 21 |
+
|
| 22 |
+
def get_vectorstore():
|
| 23 |
+
### Xử lý tất cả các tài liệu và nhét vào database
|
| 24 |
+
folder_path = "syllabus_nct_word_format/"
|
| 25 |
+
docx_files = list_docx_files(folder_path)
|
| 26 |
+
|
| 27 |
+
all_splits = [] # Khởi tạo danh sách lưu kết quả
|
| 28 |
+
for i, file_path in enumerate(tqdm(docx_files, desc="Đang xử lý", unit="file")):
|
| 29 |
+
output_json_path = f"output_{i}.json"
|
| 30 |
+
splits = get_splits(file_path, output_json_path)
|
| 31 |
+
all_splits += splits
|
| 32 |
+
|
| 33 |
+
# Xử lý FAQ
|
| 34 |
+
FAQ_path = "syllabus_nct_word_format/FAQ.json"
|
| 35 |
+
FAQ_splits = get_json_splits_only(FAQ_path)
|
| 36 |
+
all_splits += FAQ_splits
|
| 37 |
+
|
| 38 |
+
# Lưu vào vectorstore với nhúng từ Google GenAI
|
| 39 |
+
embedding = GoogleGenerativeAIEmbeddings(model="models/text-embedding-004")
|
| 40 |
+
vectorstore = FAISS.from_documents(documents=all_splits, embedding=embedding)
|
| 41 |
+
|
| 42 |
+
return vectorstore
|
| 43 |
+
|
| 44 |
+
###
|
| 45 |
|
| 46 |
async def get_urls_splits(url='https://nct.neu.edu.vn/', char='https://nct.neu.edu.vn/'):
|
| 47 |
reqs = requests.get(url)
|