summaryrefslogtreecommitdiff
path: root/rag/parser/pdf.py
diff options
context:
space:
mode:
Diffstat (limited to 'rag/parser/pdf.py')
-rw-r--r--rag/parser/pdf.py10
1 files changed, 4 insertions, 6 deletions
diff --git a/rag/parser/pdf.py b/rag/parser/pdf.py
index cb63fc4..a5e4665 100644
--- a/rag/parser/pdf.py
+++ b/rag/parser/pdf.py
@@ -1,16 +1,14 @@
+import os
from pathlib import Path
-from langchain_community.document_loaders import PyPDFLoader
-from langchain.text_splitter import RecursiveCharacterTextSplitter
-CHUNK_SIZE = 1024
-CHUNK_OVERLAP = 256
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain_community.document_loaders import PyPDFLoader
def parser(filepath: Path):
content = PyPDFLoader(filepath).load()
- print(content)
splitter = RecursiveCharacterTextSplitter(
- chunk_size=CHUNK_SIZE, chunk_overlap=CHUNK_OVERLAP
+ chunk_size=os.environ["CHUNK_SIZE"], chunk_overlap=os.environ["CHUNK_OVERLAP"]
)
chunks = splitter.split_documents(content)
return chunks