Skip to content

render.md

Latest commit

 

History

History
83 lines (62 loc) · 2.52 KB

render.md

File metadata and controls

83 lines (62 loc) · 2.52 KB

Render Module

The Render module (src/journal_utilities/render/) transforms YouTube playlists into structured course directories with transcript-based module.md files.

Architecture

graph LR
    P[Playlist Manifest] --> S[scaffold_course]
    T[Transcripts Dir] --> S
    S --> C[Course Directory]
    C --> M[module.md files]
    C --> J[course.json metadata]
Loading

Components

renderer.py

Function Description
slugify(title, max_length=60) Convert titles to filesystem-safe slugs (lowercase, hyphens, truncation on word boundaries)
format_duration(seconds) Format seconds to H:MM:SS or M:SS strings
format_upload_date(upload_date) Convert YYYYMMDD to YYYY-MM-DD
render_module_md(...) Generate module.md content with YAML metadata, YouTube link, and transcript text
scaffold_course(...) Create numbered module directories with module.md files from a playlist

Course Directory Structure

data/courses/{course-slug}/
├── course.json          # Playlist metadata (title, slug, video_count, last_updated)
├── 01_video-title/
│   └── module.md        # Transcript with YouTube link and metadata
├── 02_another-video/
│   └── module.md
└── ...

module.md Format

Each generated module contains:

  • Title (H1 heading)
  • Source link to the original YouTube video
  • Playlist name (if available)
  • Duration and upload date
  • Transcript method (e.g., auto_caption)
  • Full transcript text

Usage

Via Script

# Scaffold courses from playlists
python scripts/scaffold_youtube_courses.py

# Enumerate playlists only (no downloads)
python scripts/scaffold_youtube_courses.py --enumerate-only

# Full pipeline: enumerate, download transcripts, scaffold
python scripts/scaffold_youtube_courses.py --download --max-playlists 3

Library

from journal_utilities.render.renderer import scaffold_course, slugify

result = scaffold_course(
    course_slug="active-inference-livestream",
    videos=[{"id": "abc123", "title": "Livestream #1", "duration": 3600}],
    transcript_dir=Path("data/output"),
    courses_dir=Path("data/courses"),
    playlist_title="Active Inference Livestream",
)
print(f"Created {result['created']} modules")

Configuration

The render module does not require configuration beyond the input directories. It is typically invoked via scripts/scaffold_youtube_courses.py, which accepts CLI arguments for transcript and course directories.