Spaces:
Sleeping
Sleeping
File size: 2,197 Bytes
63590e6 02ce710 63590e6 02ce710 af92b4d 02ce710 d287ce2 af92b4d d287ce2 63590e6 02ce710 af92b4d 63590e6 af92b4d 63590e6 d287ce2 63590e6 af92b4d 63590e6 af92b4d 63590e6 af92b4d | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 | import sys
PDF_FOLDER = "data/pdf_text"
IMAGE_FOLDER = "data/images"
if __name__ == "__main__":
args = sys.argv[1:]
mode = args[0] if args else "all"
contextual = "--contextual" in args
if mode == "all":
import subprocess
extra = ["--contextual"] if contextual else []
subprocess.run([sys.executable, __file__, "pdf"] + extra, check=True)
subprocess.run([sys.executable, __file__, "images"] + extra, check=True)
elif mode == "pdf":
from src.parsers.router import parse_folder as parse_folder_pdf
from src.chunker import chunk_pages
from src.vector_store import add_chunks
label = "CONTEXTUAL" if contextual else "STANDARD"
print(f"=== INGEST PDF ({label}) ===")
pdf_pages = parse_folder_pdf(PDF_FOLDER, extensions={".pdf"})
print(f"Tổng: {len(pdf_pages)} đoạn từ PDF")
chunks = chunk_pages(pdf_pages)
print(f"Tổng: {len(chunks)} chunks")
if contextual and chunks:
from src.contextualizer import add_context_to_chunks
print("Đang sinh context cho từng chunk...")
chunks = add_context_to_chunks(chunks, pdf_pages)
print("Context xong!")
if chunks:
add_chunks(chunks)
print("PDF xong!\n")
elif mode == "images":
from src.parsers.router import parse_folder as parse_folder_img
from src.chunker import chunk_pages
from src.vector_store import add_chunks
label = "CONTEXTUAL" if contextual else "STANDARD"
print(f"=== INGEST ẢNH ({label}) ===")
image_pages = parse_folder_img(IMAGE_FOLDER, extensions={".jpg", ".jpeg", ".png"})
print(f"Tổng: {len(image_pages)} đoạn từ ảnh")
chunks = chunk_pages(image_pages)
print(f"Tổng: {len(chunks)} chunks")
if contextual and chunks:
from src.contextualizer import add_context_to_chunks
print("Đang sinh context cho từng chunk...")
chunks = add_context_to_chunks(chunks, image_pages)
print("Context xong!")
if chunks:
add_chunks(chunks)
print("Ảnh xong!")
|