Update app.py

app.py

@@ -2,6 +2,8 @@ import streamlit as st
 import pandas as pd
 from openai import OpenAI
 import os 
+import json
+IMPORT pypdf
 import subprocess
 TOKEN=os.getenv('HF_TOKEN')
 subprocess.run(["huggingface-cli", "login", "--token", TOKEN, "--add-to-git-credential"])
@@ -10,11 +12,6 @@ OPENAI_API_KEY = os.getenv("OPENAI_API")
 client = OpenAI(api_key=OPENAI_API_KEY) #INSERT KEY INSODE HE QUOTES IN THE BRACKET
 from docx import Document
 
-# Function to extract text from a .docx file
-def extract_text_from_docx(file):
-    doc = Document(file)
-    text = "\n".join([para.text for para in doc.paragraphs])
-    return text.strip()
 
 # Function to parse the feedback into rubric components
 def parse_feedback(feedback):
@@ -44,7 +41,7 @@ def parse_feedback(feedback):
     return scores
 
 # Function to grade the essay using GPT-4
-def grade_essay(essay, guided_data, topic, rubric):
+def grade_essay(essay, guided_data, rubric):
     # Sample prompt for grading using GPT-4
     prompt = f"""
     You are an consultant that grades marketing and business proposal based on a provided rubric, ensuring an unbiased evaluation while considering clarity, originality, organization, and depth of analysis. Advise in Vietnamse, only use English for buzzwords.
@@ -72,7 +69,13 @@ def grade_essay(essay, guided_data, topic, rubric):
         {"role": "user", "content": prompt}
     ])
     return response.choices[0].message.content
-
+def read_pdf(pdf_reader):
+    for page in pdf_reader.pages:
+        page_text = page.extract_text()
+        if page_text:
+            all_text += page_text + "\n"
+            return all_text
+                    
 # Function to export results to CSV
 def export_to_csv(data):
     df = pd.DataFrame(data)
@@ -99,57 +102,38 @@ def main():
         st.session_state.results = []
 
     # File uploader for example graded essays (DOCX)
-    example_files = st.file_uploader("Upload 10 example graded essays (DOCX)", type=["docx"], accept_multiple_files=True)
+    # example_files = st.file_uploader("Upload 10 example graded essays (DOCX)", type=["docx"], accept_multiple_files=True)
+        for filename in os.listdir("data"):
+            if filename.lower().endswith(".pdf"):
+                pdf_path = os.path.join(pdf_directory, filename)
+                with open(pdf_path, "rb") as pdf_file:
+                    pdf_reader = PyPDF2.PdfReader(pdf_file)
+                    example_files = read_pdf(pdf_reader)
+            
+         
 
     # File uploader for corresponding scores (DOCX)
-    scores_file = st.file_uploader("Upload the DOCX file containing corresponding scores", type=["xlsx"])
-
+    # scores_file = st.file_uploader("Upload the json file containing corresponding scores", type=["xlsx"])
+        # Open and read the JSON file with utf-8 encoding
+        with open('abs.json', 'r', encoding='utf-8') as file:
+            scores_file = json.load(file)
+        
     # File uploader for new essays to be graded (DOCX)
-    new_files = st.file_uploader("Upload DOCX files with essays to be graded", type=["docx"], accept_multiple_files=True)
-
+    pdf_file = st.file_uploader("Upload proposal to be graded", type=["pdf"], accept_multiple_files=True)
+    pdf_reader = PyPDF2.PdfReader(pdf_file)
+    new_file = read_pdf(pdf_reader)
     # Grading button
     if st.button("Grade Essays"):
-        if example_files and scores_file and new_files:
-            # Extract scores from the scores file
-            scores_text = extract_text_from_docx(scores_file)
-            scores_lines = scores_text.splitlines()
-
-            # Create a dictionary to match scores to participant names
-            scores_dict = {}
-            for line in scores_lines:
-                if ':' in line:  # Assuming the format is "Participant Name: Score"
-                    name, score = line.split(':', 1)
-                    scores_dict[name.strip()] = score.strip()
-
-            # Prepare guided data from example graded essays
-            guided_data = {}
-            for example_file in example_files:
-                essay_text = extract_text_from_docx(example_file)
-                participant_name = os.path.splitext(example_file.name)[0]  # Assuming name is file name
-                if participant_name in scores_dict:
-                    guided_data[participant_name] = {
-                        'essay': essay_text,
-                        'score': scores_dict[participant_name]
-                    }
-
-            # Combine guided essays with their scores
-            guided_data_combined = "\n".join([f"{name}: {data['essay']} (Score: {data['score']})" for name, data in guided_data.items()])
-
-            # Process each new essay
-            for new_file in new_files:
-                new_essay = extract_text_from_docx(new_file)
-                new_participant_name = os.path.splitext(new_file.name)[0]  # Assuming name is file name
-                st.write(f"Grading essay for: {new_participant_name}")
-
+        if example_files and scores_file and new_file:
+           
                 # Grading the new essay using the provided rubric and example graded essays
-                result = grade_essay(new_essay, guided_data_combined, rubric)
+                result = grade_essay(new_file, example_files, rubric)
 
                 # Parse feedback into rubric components
                 parsed_scores = parse_feedback(result)
 
                 # Store results in session state
                 st.session_state.results.append({
-                    'Participant Name': new_participant_name,
                     'Essay File': new_file.name,
                     **parsed_scores,
                     'Feedback': result,