opening files as temp pdfs

app.py

@@ -11,7 +11,7 @@ def start_haystack():
         clean_empty_lines=True,
         clean_whitespace=True,
         clean_header_footer=True,
-        split_by="passage",
+        split_by="word",
         split_length=200,
         split_respect_sentence_boundary=True,
     )
@@ -23,12 +23,11 @@ def pdf_to_document_store(pdf_files):
     converter = PDFToTextConverter(remove_numeric_tables=True, valid_languages=["en"])
     documents = []
     for pdf in pdf_files:
-        bytes_data = pdf.read()
-        st.write(pdf)
-        doc = converter.convert(file_path=pdf.name, meta=None)[0]
-        st.write(doc)
-        preprocessed_doc=preprocessor.process([doc])
-        documents.append(preprocessed_doc)
+        with open("temp-path.pdf", 'wb') as temp_file:
+            temp_file.write(pdf)
+            doc = converter.convert(file_path="temp-path.pdf", meta=None)[0]
+            preprocessed_doc=preprocessor.process([doc])
+            documents.append(preprocessed_doc)
     document_store.write_documents(documents)
     st.write('Document count: ', document_store.get_document_count())