vector_database.py

from langchain_community.document_loaders import PDFPlumberLoader
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain_ollama import OllamaEmbeddings
from langchain_community.vectorstores import FAISS
import os

pdfs_directory = 'pdfs/'

def upload_pdf(file):
    """Uploads a PDF file to the pdfs/ directory."""
    os.makedirs(pdfs_directory, exist_ok=True)  # Ensure directory exists
    file_path = os.path.join(pdfs_directory, file.name)
    
    with open(file_path, "wb") as f:
        f.write(file.getbuffer())
    
    return file_path  # Returning path for further processing

def load_pdf(file_path):
    """Loads a PDF using PDFPlumberLoader."""
    loader = PDFPlumberLoader(file_path)
    documents = loader.load()
    return documents  # Return loaded documents
file_path = "ilovepdf_merged.pdf"
documents = load_pdf(file_path)
def create_chunks(documents):
    text_splitter = RecursiveCharacterTextSplitter(
        chunk_size =1000,
        chunk_overlap = 200,
        add_start_index = True
    )
    text_chunks= text_splitter.split_documents(documents)
    return text_chunks
text_chunks = create_chunks(documents)
#step3 : setup embeeding model (use deepseek r1 with ollama)
olama_model_name= "deepseek-r1:1.5b"
def get_embeeding_model(olama_model_name):
    embeeding = OllamaEmbeddings(model= olama_model_name)
    return embeeding

FAISS_DB_PATH= "vectorstore/db_faiss"
faiss_db =FAISS.from_documents(text_chunks,get_embeeding_model(olama_model_name))
faiss_db.save_local(FAISS_DB_PATH)