sobstel / calmbook Goto Github PK

Refactor current 'dirty' scrapper, try to extract logic for different post types - text, image(s), link, video (any other types?)

Add support for lazy loading more posts.

When there's a pinned post it's always at the top. We should sort posts by date at our end.

TODO

• routing
• move layout (header, footer) to partials (using pug includes)
• separate template for page
• ensure there's atom feed only for main (index) page

Goal: there should be one and only one page only.

Solution: use fragment (#) linking.

eg. https://www.facebook.com/FacebookPolska/

• It should use main layout template.
• It should explain how to create a page using Facebook page url (path).
• It should display error message (if there's an error present).
• It should be served for "/" too (all the same, but without error message).

Current state

We use official custom search API, but it has low free limit (up to 100 queries daily).

Expected outcome

Get search results cheaper (free?), eg use some nasty scraper or so.