# Blog Platform

The store ships with a full **content platform** — not a bolted-on news page,
but an editorial system with drafts, review, scheduling, and SEO baked in. It
is how the store does content marketing.

## Authoring

- A **rich-text editor** built on Action Text and the Lexxy editor, so authors
  format posts without touching HTML.
- **Image handling** via `BlogPostImageService`, with Unsplash integration for
  sourcing imagery.
- **Duplication** — `BlogPostDuplicator` clones a post as a starting point for
  a new one.

## Editorial workflow

A post moves through real states rather than just draft/published:

```
draft  ──▶  pending review  ──▶  scheduled  ──▶  published
```

`BlogPostPublisherService` owns the transitions, and a scheduled post is
released automatically by a background job at its publish time. Posts can also
be marked **featured** for prominent placement.

## SEO, by default

SEO is not an afterthought — it is a concern mixed into the model:

- **`BlogPost::Seoable`** — slugs, meta tags, Open Graph cards, and JSON-LD
  structured data so posts present well in search results and social unfurls.
- **`BlogPost::TocGenerator`** — builds a table of contents from the post's
  headings.
- **`BlogPost::ContentFormattable`** — normalizes rendered content.

## Reading experience

The blog index uses **cursor-based pagination** to drive infinite scroll —
the same pagination engine described in [Scaling](/architecture/scaling/) — so
long archives load smoothly without page-number drift.

Rendered posts are cached via `BlogPostCacheService` and invalidated on edit.

## Key files

| Concern | Files |
|---------|-------|
| Model | `BlogPost`, `BlogPostTag`, concerns `Seoable`, `TocGenerator`, `ContentFormattable` |
| Services | `BlogPostPublisherService`, `BlogPostImageService`, `BlogPostCacheService`, `BlogPostDuplicator` |
| UI | `app/components/blog_post/`, `CursorPaginator` |