app.py

import gradio as gr
import earthview as ev
from PIL import Image
import numpy as np
import random
import os
import json
import utils
from pandas import DataFrame

# --- Configuration ---
DATASET_SUBSET = "satellogic"
NUM_SAMPLES_TO_LABEL = 100  # You can adjust this
LABELED_DATA_FILE = "labeled_data.json"
DISPLAY_N_COOL = 5  # How many "cool" examples to display alongside each new image.
SAMPLE_SEED = 10  # The seed to use when sampling the dataset for the demo.
BATCH_SIZE = 10

# --- Load Dataset ---
dataset = ev.load_dataset(DATASET_SUBSET, shards=[SAMPLE_SEED])
data_iter = iter(dataset)

# --- Load Labeled Data (if it exists) ---
def load_labeled_data():
    if os.path.exists(LABELED_DATA_FILE):
        with open(LABELED_DATA_FILE, "r") as f:
            return json.load(f)
    else:
        return []

labeled_data = load_labeled_data()

# --- Get Next Sample for Labeling ---
def get_next_sample():
    global data_iter
    try:
        sample = next(data_iter)
        sample = ev.item_to_images(DATASET_SUBSET, sample)
        return sample
    except StopIteration:
        print("No more samples in the dataset.")
        return None

def get_images(batch_size, state):
    subset = state["subset"]
    images = []
    metadatas = []
    for _ in range(batch_size):
      sample = get_next_sample()
      if sample is None:
        break

      image = sample["rgb"][0]
      metadata = sample["metadata"]
      metadata["map"] = f'<a href="{utils.get_google_map_link(sample, subset)}" target="_blank">🧭</a>'

      images.append(image)
      metadatas.append(metadata)

    return images, DataFrame(metadatas)

# --- Save Labeled Data ---
def save_labeled_data(label, state):
    global labeled_data
    sample = state["sample"]
    if sample is None:
        return "No image to label", None, DataFrame()

    image = sample["rgb"][0]  # Get the PIL Image object

    image_bytes = image.convert("RGB").tobytes()
    labeled_data.append({
        "image": image_bytes,
        "metadata": sample["metadata"],
        "label": label
    })

    with open(LABELED_DATA_FILE, "w") as f:
        json.dump(labeled_data, f)

    new_sample = get_next_sample()
    if new_sample is None:
      state["sample"] = None
      return "Dataset exhausted.", None, DataFrame()

    state["sample"] = new_sample

    new_image = new_sample["rgb"][0]
    new_metadata = new_sample["metadata"]
    new_metadata["map"] = f'<a href="{utils.get_google_map_link(new_sample, DATASET_SUBSET)}" target="_blank">🧭</a>'
    
    return "", new_image, DataFrame([new_metadata])

# --- Gradio Interface ---
# --- Labeling UI ---
def labeling_ui():
    
    state = gr.State({"sample": None, "subset": DATASET_SUBSET})

    with gr.Row():
        with gr.Column():
            gallery = gr.Gallery(label="Satellite Image", interactive=False, columns=1, object_fit="scale-down")

            with gr.Row():
                cool_button = gr.Button("Cool")
                not_cool_button = gr.Button("Not Cool")

            table = gr.DataFrame(datatype="html")

            def initialize_labeling_ui():
              sample = get_next_sample()
              
              image, metadata = get_images(1, {"sample": None, "subset": DATASET_SUBSET})

              return sample, image, metadata

            initial_sample, initial_image, initial_metadata = initialize_labeling_ui()
            gallery.value = initial_image
            table.value = initial_metadata
            state.value["sample"] = initial_sample

            # Handle button clicks
            cool_button.click(
                fn=lambda label, state: save_labeled_data(label, state),
                inputs=[gr.Textbox(visible=False, value="cool"), state],
                outputs=[gr.Textbox(label="Debug"), gallery, table]
            )
            not_cool_button.click(
                fn=lambda label, state: save_labeled_data(label, state),
                inputs=[gr.Textbox(visible=False, value="not cool"), state],
                outputs=[gr.Textbox(label="Debug"), gallery, table]
            )

# --- Display UI ---
def display_ui():

    def get_random_cool_images(n):
        cool_samples = [d for d in labeled_data if d["label"] == "cool"]
        return [Image.frombytes("RGB", (384, 384), s["image"]) for s in cool_samples] if len(cool_samples) >= n else []

    def get_new_unlabeled_image():
        global data_iter
        try:
            sample = next(data_iter)
            sample = ev.item_to_images(DATASET_SUBSET, sample)
            return sample["rgb"][0], json.dumps(sample["metadata"]["bounds"])
        except StopIteration:
            print("No more samples in the dataset.")
            return None, None

    def refresh_display():
        new_image, new_metadata = get_new_unlabeled_image()
        cool_images = get_random_cool_images(DISPLAY_N_COOL)
        if new_image is None:
            return "No more samples", None, []
        return "", new_image, cool_images

    with gr.Row():
        new_image_component = gr.Image(label="New Image", type="pil")
        metadata_display = gr.Textbox(label="Metadata (Bounds)")

    with gr.Row():
        cool_images_gallery = gr.Gallery(label="Cool Examples", value=[], columns=DISPLAY_N_COOL)

    refresh_button = gr.Button("Refresh")
    refresh_button.click(
        fn=refresh_display,
        inputs=[],
        outputs=[gr.Textbox(label="Debug"), new_image_component, cool_images_gallery]
    )

    def initialize_display_ui():
        debug, image, gallery = refresh_display()
        return debug, image, gallery

    debug, initial_image, initial_gallery = initialize_display_ui()
    new_image_component.value = initial_image
    cool_images_gallery.value = initial_gallery

# --- Main Interface ---
with gr.Blocks() as demo:
    gr.Markdown("# TerraNomaly")
    with gr.Tabs():
        with gr.TabItem("Labeling"):
            labeling_ui()
        with gr.TabItem("Display"):
            display_ui()

demo.launch(debug=True)