from PyPDF2 import PdfReader from openai import AzureOpenAI import gradio as gr import os class IncompleteSentenceFinder: """ This class finds and displays incomplete sentences in a PDF document using OpenAI's GPT-3. Args: api_key (str): Your OpenAI API key. """ def __init__(self): """ Initialize the IncompleteSentenceFinder with the PDF file and OpenAI API key. Args: api_key (str): Your OpenAI API key. """ # openai.api_type = os.getenv['api_type'] # openai.api_base = os.getenv['api_base'] # openai.api_version = os.getenv['api_version'] # openai.api_key = os.getenv['api_key'] pass def _check_incomplete_sentence(self, text: str) -> str: """ Use OpenAI's GPT-3 to identify incomplete sentences in the given text. Args: text (str): Text to check for incomplete sentences. Returns: str: Incomplete sentences identified by GPT-3. """ client = AzureOpenAI(api_key=os.getenv("AZURE_OPENAI_KEY"), api_version="2023-07-01-preview", azure_endpoint = os.getenv("AZURE_OPENAI_ENDPOINT") ) conversation = [ {"role": "system", "content": "You are a helpful incomplete sentences finder"}, {"role": "user", "content": f"""list out the incomplete sentences in the following text: {text}"""} ] # Call OpenAI GPT-3.5-turbo chat_completion = client.chat.completions.create( model = "GPT-3", messages = conversation, max_tokens=1000, temperature=0 ) response = chat_completion.choices[0].message.content return response def get_incomplete_sentence(self,pdf_file_path) -> str: """ Extract text from the PDF document and find incomplete sentences. Returns: str: Incomplete sentences identified by GPT-3. """ try: # Open the multi-page PDF using PdfReaderer pdf = PdfReader(pdf_file_path.name) incomplete_text = "" # Extract text from each page and pass it to the process_text function for page_number in range(len(pdf.pages)): # Extract text from the page page = pdf.pages[page_number] text = page.extract_text() incomplete_text += self._check_incomplete_sentence(text) return incomplete_text except Exception as e: print(f"An error occurred: {str(e)}") def file_output_fnn(self,file_path): file_path = file_path.name return file_path def gradio_interface(self): with gr.Blocks(css="style.css",theme='xiaobaiyuan/theme_brief') as demo: with gr.Row(elem_id = "col-container",scale=0.80): with gr.Column(elem_id = "col-container",scale=0.80): file1 = gr.File(label="File",elem_classes="filenameshow") with gr.Column(elem_id = "col-container",scale=0.20): upload_button1 = gr.UploadButton( "Browse File",file_types=[".txt", ".pdf", ".doc", ".docx",".json",".csv"], elem_classes="uploadbutton") incomplete_sentence_btn = gr.Button("Get Headings",elem_classes="uploadbutton") with gr.Row(elem_id = "col-container",scale=0.60): headings = gr.Textbox(label = "Headings") upload_button1.upload(self.file_output_fnn,upload_button1,file1) incomplete_sentence_btn.click(self.get_incomplete_sentence,upload_button1,headings)