import gradio as gr
import pandas as pd
from datetime import datetime
import os
import plotly.graph_objects as go
from transformers import pipeline

# Load the image classification model
classifier = pipeline("image-classification", model="google/vit-base-patch16-224")

# CSV file to store reports
CSV_FILE = "eco_reports.csv"

# Load existing reports or create a new DataFrame
if os.path.exists(CSV_FILE):
    df = pd.read_csv(CSV_FILE)
else:
    df = pd.DataFrame(columns=[
        "timestamp", "country", "city", "area", "sdg_category", "description", "image_path", "latitude", "longitude"
    ])

# Submit a new report
def submit_report(country, city, area, sdg_category, description, image, latitude, longitude):
    timestamp = datetime.now().isoformat()
    image_path = image if image else ""

    new_report = {
        "timestamp": timestamp,
        "country": country,
        "city": city,
        "area": area,
        "sdg_category": sdg_category,
        "description": description,
        "image_path": image_path,
        "latitude": latitude,
        "longitude": longitude
    }

    global df
    df = pd.concat([df, pd.DataFrame([new_report])], ignore_index=True)
    df.to_csv(CSV_FILE, index=False)

    return "✅ Your report has been submitted. Thank you!"

# Filter reports by SDG or city
def filter_reports(sdg_category, city):
    filtered_df = df.copy()
    if sdg_category != "All":
        filtered_df = filtered_df[filtered_df["sdg_category"] == sdg_category]
    if city != "All":
        filtered_df = filtered_df[filtered_df["city"] == city]
    return [
        gr.Image(value=row["image_path"], label=f"{row['city']} | {row['sdg_category']} | {row['description']}")
        for _, row in filtered_df.iterrows()
    ]

# Generate a map of all reports
def generate_map():
    if df.empty or df["latitude"].isnull().all():
        return go.Figure()

    fig = go.Figure(go.Scattermapbox(
        lat=df["latitude"],
        lon=df["longitude"],
        mode="markers",
        marker=go.scattermapbox.Marker(size=9, color='green'),
        text=df["description"],
        hoverinfo="text"
    ))

    fig.update_layout(
        mapbox_style="open-street-map",
        mapbox_zoom=1.5,
        mapbox_center={"lat": 20, "lon": 0},
        margin={"l":0, "r":0, "t":0, "b":0}
    )

    return fig

# Image classification
def classify_image(image):
    predictions = classifier(image)
    return predictions[0]["label"]

# --- Gradio Interfaces ---

# Tab 1: Submit Report
report_interface = gr.Interface(
    fn=submit_report,
    inputs=[
        gr.Textbox(label="🌍 Country", placeholder="e.g. Kenya"),
        gr.Textbox(label="🌆 City", placeholder="e.g. Nairobi"),
        gr.Textbox(label="🏘️ Area", placeholder="e.g. Westlands"),
        gr.Dropdown(label="📊 SDG Category", choices=[
            "SDG 13: Climate Action", "SDG 11: Sustainable Cities", "SDG 15: Life on Land"
        ]),
        gr.Textbox(label="📝 Description", lines=3, placeholder="Describe the environmental issue..."),
        gr.Image(type="filepath", label="📷 Upload a Photo"),
        gr.Number(label="Latitude"),
        gr.Number(label="Longitude")
    ],
    outputs="text",
    title="📩 Submit an Environmental Report",
    description="Help fight climate change by reporting visible environmental issues around you."
)

# Tab 2: View Reports
view_interface = gr.Interface(
    fn=filter_reports,
    inputs=[
        gr.Dropdown(choices=["All", "SDG 13: Climate Action", "SDG 11: Sustainable Cities", "SDG 15: Life on Land"], label="Filter by SDG"),
        gr.Dropdown(choices=["All"] + sorted(df["city"].dropna().unique().tolist()), label="Filter by City")
    ],
    outputs=gr.Gallery(label="Filtered Reports"),
    title="📊 View Community Reports",
    description="Browse recent reports submitted by users."
)

# Tab 3: Map of Reports
map_interface = gr.Interface(
    fn=generate_map,
    inputs=[],
    outputs=gr.Plot(),
    title="🗺️ Map of Reports",
    description="See where environmental issues are being reported."
)

# Tab 4: AI Image Tagging
ai_interface = gr.Interface(
    fn=classify_image,
    inputs=gr.Image(type="pil", label="Upload an Image"),
    outputs=gr.Textbox(label="Predicted Category"),
    title="🤖 AI Tagging of Images",
    description="Use AI to classify the environmental issue in the photo."
)

# Launch the app with all tabs
app = gr.TabbedInterface(
    [report_interface, view_interface, map_interface, ai_interface],
    tab_names=["Submit Report", "View Reports", "Map View", "AI Tagging"]
)

app.launch()