Podcrastinator

Recently I fell in love with Made by Google Podcast.

I like listening while walking.
I like transcripts.
I like being able to tap a sentence and jump back to that exact moment.

What I do not like is manually downloading, sorting, and uploading podcast episodes one by one.

So I made Podcrastinator — a tiny tool that helps me upload one episode at a time, slowly, and listen at my own pace.

---

🌟 What it does

Podcrastinator has evolved into a local Web Console built to track, manage, download, and automatically transport RSS podcast episodes into private web platforms that lack public APIs.

• Manage Subscriptions: Add and track multiple RSS feeds from a single console.
• Download Media Locally: Fetch MP3 files and cover images reliably into a neatly organized folder structure.
• Automated Web Uploading: Leverages browser automation (Playwright) to emulate human actions—filling titles, expanding panels, waiting for dynamic uploads, and checking boxes.
• Track Progress: A persistent async state machine tracks every episode visually (Pending, Downloading, Uploaded, etc.).

🏗 Tech Stack

• Backend: FastAPI (Python 3) handling background async tasks.
• Frontend: Jinja2 Templates, HTMX (for fast, reactive UI interactions without writing JS), and Pico CSS (for a pure, clean, native-feeling UI).
• Automation: Playwright (Python).
• Storage: Directory & JSON-based tracking (metadata.json). Highly transparent and portable.

🚀 Setup & Running

1. Install Dependencies

   pip install -r requirements.txt
   playwright install chromium

2. Configure Platform Selectors Copy .env.example to .env and map the CSS selectors corresponding to the upload form of your target platform.

3. Start the Web Console

   python app.py

   Open http://127.0.0.1:8000 in your browser. Add your first podcast channel, sync the latest feed, and hit "Download" or "Upload" to watch the magic happen!

📂 Data Structure

Locally downloaded episodes are kept meticulously clean:

data/
└── podcasts/
    └── [podcast_id]/
        └── episodes/
            └── [episode_id]/
                ├── metadata.json     # The source of truth state tracker
                ├── description.txt
                ├── cover.jpg
                └── media.mp3