Initial commit for Hugging Face Space

README.md

@@ -12,3 +12,25 @@ short_description: doc gen
 ---
 
 Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# Code to Project Document Generator
+
+This is a Streamlit app that takes code files and a heading structure, and generates:
+- A DOCX project report
+- A PPTX presentation
+
+It uses lightweight models like T5 and tools like python-docx and pptx. Deployable on Hugging Face Spaces!
+
+## Usage
+1. Upload your code files (Python or React)
+2. Upload a `headings.txt` file with your report headings
+3. Click Generate
+
+## Output
+- `project_report.docx`
+- `project_presentation.pptx`
+
+## Run Locally
+```bash
+pip install -r requirements.txt
+streamlit run app.py
+```

app.py

@@ -0,0 +1,29 @@
+import streamlit as st
+from utils.summarizer import summarize_code
+from utils.doc_generator import generate_document, generate_pptx
+import os
+
+st.title("📄 Code to Project Document Generator")
+st.write("Upload your code files and a headings.txt file.")
+
+uploaded_files = st.file_uploader("Upload code files (Python/React)", accept_multiple_files=True)
+headings_file = st.file_uploader("Upload headings.txt", type="txt")
+
+if st.button("Generate Document") and uploaded_files and headings_file:
+    with open("headings.txt", "wb") as f:
+        f.write(headings_file.read())
+
+    code_dir = "uploaded_code"
+    os.makedirs(code_dir, exist_ok=True)
+
+    for file in uploaded_files:
+        with open(os.path.join(code_dir, file.name), "wb") as f:
+            f.write(file.read())
+
+    sections = summarize_code(code_dir, "headings.txt")
+    generate_document(sections)
+    generate_pptx(sections)
+
+    st.success("Documents generated!")
+    st.download_button("Download DOCX", data=open("project_report.docx", "rb"), file_name="project_report.docx")
+    st.download_button("Download PPTX", data=open("project_presentation.pptx", "rb"), file_name="project_presentation.pptx")

headings.txt

@@ -0,0 +1,27 @@
+ABSTRACT
+Introduction
+Python
+Machine Learning
+Deep Learning
+Image Classification
+Working of Image Classification
+Applications of Image Classification
+Architecture
+Data Flow Diagram
+Image Classification Techniques
+Aim of the Project
+Scope
+System Requirements
+Hardware Requirements
+Software Requirements
+Setup Instructions
+Algorithms
+Performance Evaluation
+Comparison of Base Line Models
+Error Analysis
+Methodology
+Results
+Discussion
+CONCLUSION
+REFERENCES
+FUTURE WORK

requirements.txt

@@ -0,0 +1,7 @@
+streamlit
+transformers
+sentencepiece
+docx
+python-pptx
+plantuml
+huggingface-hub

sample_code/sample.py

@@ -0,0 +1,3 @@
+# Sample Python file
+def greet(name):
+    return f"Hello, {name}!"

utils/doc_generator.py

@@ -0,0 +1,19 @@
+from docx import Document
+from pptx import Presentation
+from pptx.util import Inches
+
+def generate_document(sections):
+    doc = Document()
+    doc.add_heading("Project Report", 0)
+    for title, content in sections.items():
+        doc.add_heading(title, level=1)
+        doc.add_paragraph(content)
+    doc.save("project_report.docx")
+
+def generate_pptx(sections):
+    prs = Presentation()
+    for title, content in sections.items():
+        slide = prs.slides.add_slide(prs.slide_layouts[1])
+        slide.shapes.title.text = title
+        slide.placeholders[1].text = content
+    prs.save("project_presentation.pptx")

utils/summarizer.py

@@ -0,0 +1,20 @@
+from transformers import pipeline
+
+summarizer = pipeline("summarization", model="t5-small")
+
+def summarize_code(code_dir, headings_path):
+    sections = {}
+    with open(headings_path, "r") as hfile:
+        headings = [line.strip() for line in hfile if line.strip()]
+
+    for heading in headings:
+        combined_code = ""
+        for root, _, files in os.walk(code_dir):
+            for file in files:
+                with open(os.path.join(root, file), "r", encoding="utf-8", errors="ignore") as f:
+                    combined_code += f.read() + "\n"
+
+        summary = summarizer(combined_code[:1000], max_length=120, min_length=30, do_sample=False)[0]["summary_text"]
+        sections[heading] = summary
+
+    return sections