Upload 3 files

app.py

@@ -0,0 +1,55 @@
+import streamlit as st
+import llm
+from file_creator import Create_Doc
+
+model = llm.Model()
+unvalid_image_text = 'please upload a valid screenshot.'
+
+
+st.markdown("""
+    <style>
+    .justified-text {
+        text-align: justify;
+    }
+    </style>
+    """, unsafe_allow_html=True)
+
+
+with st.sidebar:
+    st.header("ABOUT:")
+    
+    st.caption("""
+        <div class="justified-text">
+            Document Creator is an innovative app that allows users to effortlessly convert their screenshots into Word documents. Simply upload a screenshot, and the app will generate a Word document based on the image provided, ensuring a seamless and efficient conversion process. Ideal for anyone looking to quickly turn visual content into editable text documents.
+        </div>
+        """, unsafe_allow_html=True)
+    
+    for _ in range(17):
+        st.write("") 
+    st.subheader("Build By:")
+    st.write("[Pachaiappan❤️](https://mr-vicky-01.github.io/Portfolio)")
+    st.write("contact: [Email](mailto:pachaiappan1102@gamil.com)")
+
+st.title("Document Creator")
+st.text("Updload your Screenshot to convert into a word document.")
+uploaded_file = st.file_uploader("", type=["png", "jpg", "jpeg"])
+
+if uploaded_file:
+    st.image(uploaded_file)
+    
+    button = st.button("Generate Document")
+    if button:
+        with st.spinner("Generating a Document"):
+            text = model.get_response(uploaded_file)
+        st.write(text)
+        
+        if text.lower().strip() != unvalid_image_text:
+            doc = Create_Doc()
+            doc_buffer = doc.markdown_to_word(text)
+            st.download_button(
+            label="Download",
+            data=doc_buffer,
+            file_name="output.docx",
+            mime="application/vnd.openxmlformats-officedocument.wordprocessingml.document"
+            )
+            

file_creator.py

@@ -0,0 +1,40 @@
+from docx import Document
+from docx.enum.text import WD_PARAGRAPH_ALIGNMENT
+from docx.shared import Pt
+from io import BytesIO
+import markdown
+
+class Create_Doc:
+    def __init__(self) -> None:
+        self.doc = Document()
+    
+    def markdown_to_word(self,markdown_text):
+        # Convert Markdown to HTML
+        # html = markdown.markdown(markdown_text)
+
+        # Parse the Markdown text and add formatted content to the document
+        for line in markdown_text.split('\n'):
+            if line.startswith('# '):
+                heading = line[2:]
+                p = self.doc.add_heading(heading, level=1)
+            elif line.startswith('## '):
+                heading = line[3:]
+                p = self.doc.add_heading(heading, level=2)
+            elif line.startswith('### '):
+                heading = line[4:]
+                p = self.doc.add_heading(heading, level=3)
+            elif line.startswith('- '):
+                item = line[2:]
+                p = self.doc.add_paragraph(item, style='ListBullet')
+            else:
+                p = self.doc.add_paragraph(line)
+            
+            # Adjust paragraph formatting (optional)
+            p.alignment = WD_PARAGRAPH_ALIGNMENT.LEFT
+            p.style.font.size = Pt(12)
+
+        # Save the document to a BytesIO object
+        buffer = BytesIO()
+        self.doc.save(buffer)
+        buffer.seek(0)
+        return buffer

llm.py

@@ -0,0 +1,19 @@
+import google.generativeai as genai
+from dotenv import load_dotenv
+from PIL import Image
+import os
+
+load_dotenv()
+os.environ["GOOGLE_API_KEY"] = os.getenv("GOOGLE_API_KEY")
+genai.configure(api_key=os.environ["GOOGLE_API_KEY"])
+
+class Model:
+    def __init__(self) -> None:
+        self.model = genai.GenerativeModel('gemini-1.5-flash-latest')
+        
+    def get_response(self, image):
+        prompt = """You are an intelligent document creator. Could you please extract the words from the given screenshot and provide me document text that matches exact screenshot font and look
+        important note: if the screenshot not contain any text means you must say 'please upload a valid screenshot.'"""
+        img = Image.open(image)
+        response = self.model.generate_content([prompt, img]) 
+        return response.text