Spaces:
Runtime error
Runtime error
File size: 1,861 Bytes
85252a0 | 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 61 62 63 | """
One-time (or re-runnable) ingestion script.
Loads all PDFs from the documents/ directory, chunks them, embeds the
chunks, and upserts them into Pinecone. Run this once before starting the
API, and re-run whenever documents/ changes.
Usage:
python -m scripts.ingest
"""
import logging
import sys
from pathlib import Path
from app.core.config import settings
from app.core.logging_config import configure_logging
from app.services import chunker, embedder, pdf_processor, vector_store
configure_logging()
logger = logging.getLogger(__name__)
def main() -> None:
docs_dir = Path(settings.DOCUMENTS_DIR)
logger.info(f"Starting ingestion from '{docs_dir}'...")
try:
all_pages = pdf_processor.load_all_documents(docs_dir)
except pdf_processor.PDFProcessingError as e:
logger.error(f"Ingestion aborted β PDF processing failed: {e}")
sys.exit(1)
try:
all_chunks = chunker.chunk_all_documents(
all_pages,
chunk_size_words=settings.CHUNK_SIZE_WORDS,
overlap_words=settings.CHUNK_OVERLAP_WORDS,
)
except ValueError as e:
logger.error(f"Ingestion aborted β chunking failed: {e}")
sys.exit(1)
try:
texts = [c.text for c in all_chunks]
embeddings = embedder.embed_texts(texts)
except embedder.EmbeddingError as e:
logger.error(f"Ingestion aborted β embedding failed: {e}")
sys.exit(1)
try:
vector_store.ensure_index_exists()
upserted_count = vector_store.upsert_chunks(all_chunks, embeddings)
except vector_store.VectorStoreError as e:
logger.error(f"Ingestion aborted β vector store operation failed: {e}")
sys.exit(1)
logger.info(f"β
Ingestion complete. {upserted_count} chunks indexed into Pinecone.")
if __name__ == "__main__":
main() |