app.py

import gradio as gr
from video_processing import download_video, find_scenes, analyze_scenes, extract_best_scene, cleanup_temp_files
from gradio.themes.base import Base
from gradio.themes.utils import colors, fonts, sizes
from typing import Iterable
import uuid
import os
import plotly.graph_objects as go

class CustomTheme(Base):
    def __init__(
        self,
        *,
        primary_hue: colors.Color | str = colors.orange,
        secondary_hue: colors.Color | str = colors.orange,
        neutral_hue: colors.Color | str = colors.gray,
        spacing_size: sizes.Size | str = sizes.spacing_md,
        radius_size: sizes.Size | str = sizes.radius_md,
        text_size: sizes.Size | str = sizes.text_md,
        font: fonts.Font | str | Iterable[fonts.Font | str] = (
            fonts.GoogleFont("Sora"),
            "ui-sans-serif",
            "sans-serif",
        ),
        font_mono: fonts.Font | str | Iterable[fonts.Font | str] = (
            fonts.GoogleFont("Sora"),
            "ui-monospace",
            "monospace",
        ),
    ):
        super().__init__(
            primary_hue=primary_hue,
            secondary_hue=secondary_hue,
            neutral_hue=neutral_hue,
            spacing_size=spacing_size,
            radius_size=radius_size,
            text_size=text_size,
            font=font,
            font_mono=font_mono,
        )
        super().set(
            body_background_fill="radial-gradient(circle at center, rgba(235, 87, 38, 1) 0%, rgba(235, 87, 38, 0) 70%), radial-gradient(#eb5726 1px, transparent 1px)",
            body_text_color="#282828",
            block_background_fill="#ffffff",
            block_title_text_color="#eb5726",
            block_label_text_color="#eb5726",
            button_primary_background_fill="#eb5726",
            button_primary_text_color="#ffffff",
        )

custom_theme = CustomTheme()

def save_uploaded_file(uploaded_file):
    upload_dir = "uploaded_videos"
    os.makedirs(upload_dir, exist_ok=True)
    file_path = os.path.join(upload_dir, f"{uuid.uuid4()}.mp4")
    with open(file_path, "wb") as f:
        f.write(uploaded_file)
    return file_path

def display_results(video_url, video_file, description):
    if video_url:
        video_path = download_video(video_url)
    elif video_file:
        video_path = save_uploaded_file(video_file)
    else:
        return "No video provided", None, None

    scenes = find_scenes(video_path)
    if not scenes:
        return "No scenes detected", None, None

    best_scene_times, sentiments = analyze_scenes(video_path, scenes, description)
    if not best_scene_times:
        return "No matching scene found", None, None

    final_clip = extract_best_scene(video_path, best_scene_times)
    if final_clip:
        output_dir = "output"
        os.makedirs(output_dir, exist_ok=True)
        final_clip_path = os.path.join(output_dir, f"{uuid.uuid4()}_final_clip.mp4")
        final_clip.write_videofile(final_clip_path, codec='libx264', audio_codec='aac')
        cleanup_temp_files()

        # Calculate the total sum of sentiment scores
        total_score = sum(sentiments.values())
        if total_score == 0:
            # Ensure there's no division by zero
            sentiments = {k: 0 for k in sentiments}

        # Prepare data for the radial chart
        labels = list(sentiments.keys())
        values = [v / total_score * 100 for v in sentiments.values()]  # Normalize to percentages

        # Create a polar chart
        fig = go.Figure(data=go.Scatterpolar(
            r=values,
            theta=labels,
            fill='toself'
        ))

        fig.update_layout(
            polar=dict(
                radialaxis=dict(
                    visible=True,
                    range=[0, max(values) if values else 1]
                )),
            showlegend=False
        )

        return final_clip_path, final_clip_path, fig
    else:
        return "No matching scene found", None, None
        

# Custom CSS for additional styling
css = """
body {
    background-image: radial-gradient(#e83f07 1px, transparent 1px);
    background-size: 10px 10px;
    background-repeat: repeat;
    animation: adjustHue 60s infinite linear;
}

@keyframes adjustHue {
    0% { filter: hue-rotate(0deg); }
    50% { filter: hue-rotate(180deg); }
    100% { filter: hue-rotate(0deg); }
}
#video_url {
    background-color: #ffffff;
    color: #282828;
    border: 2px solid #eb5726;
}
#description {
    background-color: #ffffff;
    color: #282828;
    border: 2px solid #eb5726;
}
#submit_button {
    background-color: #eb5726;
    color: #ffffff;
    border: 2px solid #ffffff;
}
#submit_button:hover {
    background-color: #f5986e;
    color: #ffffff;
    border: 2px solid #ffffff;
}
label[for="video_url"] {
    color: #eb5726 !important;
}
label[for="description"] {
    color: #eb5726 !important;
}
h3 {
    color: #f07102;
}
.centered-markdown {
    text-align: center;
    background-color: #ffffff;
    padding: 10px;
}
#sickstadium-title {
    font-size: 5em !important;
    font-weight: bold;
    text-transform: uppercase;
}
"""

with gr.Blocks(theme=custom_theme, css=css) as demo:
    with gr.Column():
        gr.Markdown("# **Sickstadium AI**", elem_classes="centered-markdown", elem_id="sickstadium-title")
        gr.Markdown("### Upload your videos. Find sick clips. Tell your truth.", elem_classes="centered-markdown")
        gr.Markdown("**Welcome to Sickstadium AI. Our goal is to empower content creators with the ability to tell their stories without the friction of traditional video editing software. Skip the timeline, and don't worry about your video editing skills. Upload your video, describe the clip you want, and let our AI video editor do the work for you. Get more info about the Sickstadium project at [Strongholdlabs.io](https://strongholdlabs.io/)**", elem_classes="centered-markdown")
        video_url = gr.Textbox(label="Video URL:")
        video_file = gr.File(label="Upload Video File:", type="binary")
        description = gr.Textbox(label="Describe your clip:")
        submit_button = gr.Button("Process Video", elem_id="submit_button")
        video_output = gr.Video(label="Processed Video:")
        download_output = gr.File(label="Download Processed Video:")
        sentiment_output = gr.Plot(label="Predicted User Feedback:")  # Changed from Markdown to Plot
        submit_button.click(fn=display_results, inputs=[video_url, video_file, description], outputs=[video_output, download_output, sentiment_output])

demo.launch()