working version

app.py

@@ -5,14 +5,9 @@ import os
 from openpyxl import load_workbook, Workbook
 from io import BytesIO
 import openai
+import base64
 
-def save_workbook_to_bytes(wb):
-    # Save the workbook into a BytesIO object (in memory, not on disk)
-    byte_io = BytesIO()
-    wb.save(byte_io)
-    byte_io.seek(0)  # Go to the beginning of the BytesIO buffer
-    return byte_io.getvalue()
-
+from src.utils import save_workbook_to_bytes, serve_excel_for_download, image_to_base64, download_image
 
 openai_api_key = os.getenv("OPENAI_API_KEY")
 
@@ -21,12 +16,20 @@ openai.api_key = openai_api_key
 
 st.title("AutoAssess: Student Essay Transcription and Assessment")
 
-st.title("AutoAssess")
-st.write("If you see this, the basic app is loading correctly!")
-
+st.subheader("Upload Student Essays")
 # Upload folder of images
-uploaded_files = st.file_uploader("Upload a folder of student essays (images)", type=['jpg', 'jpeg', 'png'], accept_multiple_files=True)
+uploaded_files = st.file_uploader("Upload a folder of student essays (images). Please make sure that the images are in order i.e. first image is the first page of the first student, the second image is the second page of the first student (if it exists) untill all images of the first student are uploaded which should be followed by the ordererd pages of the second student, etc.", type=['jpg', 'jpeg', 'png'], accept_multiple_files=True)
+
+# Path to your example images (you need to place them in your project directory or provide a URL)
+image_1_path = "example_essay_1.png"  # Example file path
+image_2_path = "example_essay_2.png"  # Example file path
 
+# Example images section
+st.text("If you don't have any images to upload, you can download (and reupload) the example images below.")
+
+# Download buttons for the two example essays (images)
+download_image(image_1_path, "example_essay_1.png")
+download_image(image_2_path, "example_essay_2.png")
 # replace uploaded files with files loading from directory
 # image_dir = "data/images"
 # uploaded_files = []
@@ -37,9 +40,25 @@ uploaded_files = st.file_uploader("Upload a folder of student essays (images)",
 # Text inputs for question and criteria
 # essay_question = st.text_input("Enter the essay question:")
 # grading_criteria = st.text_area("Enter grading criteria or relevant marking information:")
+st.subheader("Specify Essay Question")
+st.text("For example: What is beauty?")
+essay_question = st.text_input("Enter the essay question:")
+
+st.subheader("Specify Grading Criteria")
+st.text("For example: Please follow the german grading system where 1 is the best grade and 6 is the worst grade. An essay with a grade of 1 should be well structured, have a clear introduction, body and conclusion, be free of grammatical errors, spelling mistakes and punctuation errors, be original and creative.")
+grading_criteria = st.text_area("Enter grading criteria or relevant marking information:")
+
+
+# essay_question = "What is beauty?"
+# grading_criteria = "1. Introduction\n2. Body\n3. Conclusion\n4. Grammar\n5. Spelling\n6. Punctuation\n7. Originality\n8. Creativity"
+
+
 
-essay_question = "What is beauty?"
-grading_criteria = "1. Introduction\n2. Body\n3. Conclusion\n4. Grammar\n5. Spelling\n6. Punctuation\n7. Originality\n8. Creativity"
+# Path to the example Excel file in the repository
+st.subheader("Specify Grading Criteria")
+example_file_path ="example.xlsx"
+serve_excel_for_download(example_file_path)
+st.text("Please download the example Excel file to see the expected format. You will need to adjust the student IDs and page counts. Note that if you provide a grade and reason for a student, the system will not reassess that student's essay. Further, this will help te system make better assessments by calibrating it to your style of grading. Note that grade and reason must both be specified.")
 
 # Upload Excel file with student IDs and page count
 excel_file = st.file_uploader("Upload Excel file with student IDs and page count", type=["xlsx"])
@@ -67,8 +86,8 @@ if st.button("Process Essays"):
 
         # Display the download button
         st.download_button(
-            label="Download the Excel file",
+            label="Download the assessed Excel file",
             data=excel_file,
-            file_name="results.xlsx",
+            file_name="assessments.xlsx",
             mime="application/vnd.openxmlformats-officedocument.spreadsheetml.sheet"
 )

src/transcribe_image.py

@@ -1,6 +1,34 @@
 import base64
 from openai import OpenAI
 import os
+from PIL import Image
+import io
+
+def resize_image(image, max_size=800):
+    """Resize the image to ensure the max side length is `max_size` while maintaining aspect ratio."""
+    # Open the image using Pillow
+    img = Image.open(image)
+    
+    # Get the current width and height of the image
+    width, height = img.size
+    
+    # Resize the image if necessary
+    if width > height:
+        new_width = max_size
+        new_height = int((new_width / width) * height)
+    else:
+        new_height = max_size
+        new_width = int((new_height / height) * width)
+    
+    # Resize the image using the LANCZOS filter for high-quality rescaling
+    img = img.resize((new_width, new_height), Image.Resampling.LANCZOS)
+    
+    # Save the resized image to a BytesIO object to later encode to base64
+    img_byte_arr = io.BytesIO()
+    img.save(img_byte_arr, format='PNG')
+    img_byte_arr.seek(0)  # Rewind the BytesIO object to the beginning
+    
+    return img_byte_arr
 
 # Function to encode the image
 def encode_image(image_path):
@@ -11,7 +39,8 @@ def encode_image(image_path):
 def encode_image_from_uploaded_file(image):
     # Convert image to bytes
     assert image is not None, "No image uploaded."
-    image_bytes = image.read()
+    resized_image = resize_image(image)
+    image_bytes = resized_image.read()
     return base64.b64encode(image_bytes).decode('utf-8')
 
 def transcribe_image(image_file):

src/utils.py

@@ -0,0 +1,42 @@
+import base64
+from io import BytesIO
+import streamlit as st
+
+
+
+def save_workbook_to_bytes(wb):
+    # Save the workbook into a BytesIO object (in memory, not on disk)
+    byte_io = BytesIO()
+    wb.save(byte_io)
+    byte_io.seek(0)  # Go to the beginning of the BytesIO buffer
+    return byte_io.getvalue()
+
+# Function to serve the file for download
+def serve_excel_for_download(file_path):
+    # Open the Excel file in binary mode
+    with open(file_path, "rb") as f:
+        file_data = f.read()
+
+    # Provide the download button for the user
+    st.download_button(
+        label="Download Example Excel File",
+        data=file_data,
+        file_name="example.xlsx",
+        mime="application/vnd.openxmlformats-officedocument.spreadsheetml.sheet"
+    )
+
+    # Function to convert image to base64 so that it can be displayed
+def image_to_base64(image_path):
+    with open(image_path, "rb") as image_file:
+        return base64.b64encode(image_file.read()).decode("utf-8")
+
+# Function to allow downloading the example images
+def download_image(image_path, image_name):
+    with open(image_path, "rb") as file:
+        btn = st.download_button(
+            label=f"Download {image_name}",
+            data=file,
+            file_name=image_name,
+            mime="image/png"
+        )
+    return btn