roboflowdataClean.py

import streamlit as st
import pandas as pd
import json
from io import StringIO

st.set_page_config(
    page_title="ECGO Brand Flavor Data Extractor",
    page_icon="🍹",
    layout="wide"
)
st.title("📦 ECGO Product Data Brand + Flavor Parser ~ Roboflow")
st.markdown("Convert your COCO-exported Roboflow dataset into a clean, structured CSV with **Brand Names** and **Flavor Tags** 🍫🥤")

st.markdown("---")

uploaded_file = st.file_uploader("📁 Upload your `_annotations.coco.json` file", type="json")

if uploaded_file:
    st.markdown("✅ File uploaded successfully! Parsing...")
    data = json.load(uploaded_file)

    # Extract brand names from categories
    brand_lookup = {c['id']: c['name'] for c in data.get("categories", [])}

    # Map image IDs to filenames and user_tags (flavors)
    image_lookup = {}
    for img in data.get("images", []):
        image_lookup[img['id']] = {
            'filename': img.get('extra', {}).get('name', img['file_name']),
            'tags': img.get('extra', {}).get('user_tags', [])
        }

    # Map image IDs to brand names via annotations
    brand_map = {}
    for ann in data.get("annotations", []):
        img_id = ann.get('image_id')
        cat_id = ann.get('category_id')
        brand = brand_lookup.get(cat_id, "Unknown")
        brand_map.setdefault(img_id, set()).add(brand)

    # Build final structured data
    structured_data = []
    for img_id, info in image_lookup.items():
        if img_id not in brand_map:  # Skip if no brand label exists
            continue

        row = {
            "📸 Filename": info['filename'],
            "🏷️ Brand Name": ", ".join(sorted(brand_map.get(img_id, [])))
        }
        # Add flavor tags as Flavor 1, Flavor 2, ...
        for i, tag in enumerate(info['tags']):
            row[f"🍬 Flavor {i+1}"] = tag
        structured_data.append(row)

    df = pd.DataFrame(structured_data)
    st.success("🎉 Successfully parsed JSON and generated structured data!")

    st.markdown("### 🧾 Preview of Output Table")
    st.dataframe(df, use_container_width=True)

    csv = df.to_csv(index=False).encode('utf-8')
    st.download_button("⬇️ Download CSV", csv, file_name="brand_flavor_data.csv", mime='text/csv')
else:
    st.info("👆 Upload your COCO JSON file above to get started!")

st.markdown("---")
st.markdown("<div style='text-align: center; color: gray;'>Made with ❤️ by the ECGO Data Analysis Team</div>", unsafe_allow_html=True)