VoyageV2 / store_index.py
karwanjiru's picture
Uploaded files
11916f0
import os
import time
from src.helper import PINECONE_API_KEY, text_split, download_hugging_face_embeddings
from langchain.vectorstores import Pinecone as LangchainPinecone # Alias to avoid confusion
from dotenv import load_dotenv
from pinecone import Pinecone, ServerlessSpec
from langchain_pinecone import PineconeVectorStore
from PyPDF2 import PdfReader
# Define the load_pdf function
def load_pdf(file_path):
all_text = ""
with open(file_path, 'rb') as file:
reader = PdfReader(file)
for page in reader.pages:
all_text += page.extract_text() + "\n"
return all_text if all_text else None
# Define the text_split function
def text_split(text):
from langchain.text_splitter import CharacterTextSplitter
text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)
return text_splitter.split_text(text)
# Load environment variables if not already set
load_dotenv()
# Load and process data
pdf_file_path = "data/Okelloetal.2008TourismanalysisManka.pdf" # Update this path to your single PDF file
extracted_data = load_pdf(pdf_file_path)
if extracted_data is None:
raise ValueError("The extracted data is None. Please check the load_pdf function.")
print(f"Extracted Data: {extracted_data}")
# Split the extracted text into chunks
text_chunks = text_split(extracted_data)
if text_chunks is None:
raise ValueError("The text_chunks is None. Please check the text_split function.")
print(f"Text Chunks: {text_chunks}")
embeddings = download_hugging_face_embeddings()
if embeddings is None:
raise ValueError("The embeddings is None. Please check the download_hugging_face_embeddings function.")
print(f"Embeddings: {embeddings}")
# Ensure Pinecone API key is available
api_key = os.environ.get("PINECONE_API_KEY")
if not api_key:
raise ValueError("PINECONE_API_KEY environment variable not set.")
# Initialize Pinecone client
pc = Pinecone(api_key=api_key)
# Specify cloud and region for the serverless index
cloud = os.environ.get('PINECONE_CLOUD') or 'aws'
region = os.environ.get('PINECONE_REGION') or 'us-east-1'
spec = ServerlessSpec(cloud=cloud, region=region)
# Define the index name
index_name = "healthbot"
# Create the index if it does not exist
if index_name not in pc.list_indexes().names():
pc.create_index(
name=index_name,
dimension=384,
metric="cosine",
spec=spec
)
# Wait for the index to be ready
while not pc.describe_index(index_name).status['ready']:
time.sleep(1)
# Connect to the created index
index = pc.Index(index_name)
time.sleep(1)
# Example: Add data to the index with reduced metadata
# Create a dictionary to simulate external storage of text chunks
text_chunk_store = {}
# Function to simulate storing text chunk and returning a reference ID
def store_text_chunk(text_chunk):
chunk_id = f"chunk_{len(text_chunk_store)}"
text_chunk_store[chunk_id] = text_chunk
return chunk_id
# Add text chunks to Pinecone with reference IDs
for i, text_chunk in enumerate(text_chunks):
chunk_id = store_text_chunk(text_chunk)
embedding = embeddings.embed_query(text_chunk) # Embed the text chunk
index.upsert(
vectors=[
{
"id": f"vec_{i}",
"values": embedding,
"metadata": {"chunk_id": chunk_id} # Only store the reference ID as metadata
}
],
namespace="ns1"
)
print("Indexing completed successfully.")