Skip to content

Latest commit

 

History

History
37 lines (31 loc) · 4.78 KB

README.md

File metadata and controls

37 lines (31 loc) · 4.78 KB
Raw

An Obsidian plugin for reading rss feeds

This plugin is mostly built for my own use. Online services are good, but it is also good to be able to save important rss messages in the local machine. At present it may still look rudimentary, but its functions already quite satisfies me. It differs from the existing plugin RSS Reader in a few aspects, some of which are described in the issues I raised for that plugin some time ago.

How it works

  • It creates an icon in the left sidebar (for the mobile version the icon is located in the bottom right pop-up menu).
  • The first time you use it, you need to manually add RSS feed sources.
  • It creates nested folders feeds-reader/feeds-store/ in the current vault. All the saved items (.md files) are stored in feeds-reader, and all the feeds data (.json.frag.gzip files, i.e. gzipped json fragments.) are saved in feeds-reader/feeds-store/.
  • The top-left > is for toggling the navigation bar. This is useful when reading on a small screen.
  • Search is for searching for one or more keywords (separated by space) in the current selected feed.
  • Unread only/Show all is for toggling displaying only unread items or all items (Note: all those marked as "read" or "deleted" are considered "read").
  • Title only/Show content is for toggling whether to show the content of each item. My own experience is that most of the time a glancing over the title is enough to decide whether to read an article in detail.
  • New to old is for toggling of display orders. It cycles from New to old, Old to new, to Random.
  • Save data is for saving the feed data (after you have marking some items as read/deleted), so that your reading progress would not be accidentally lost.
  • Update all is for fetching new items for all the subscribed feeds.
  • Undo is for undoing recent mark as read and mark as deleted actions for the selected feed.
  • Add feed is for adding new feed subscriptions. The first time you use this plugin, you will need to add subscriptions with this Add. Shorter feed names are preferred, and they must be unique.
    • Be careful with the feed provider! While sanitization is performed for the feed contents using the sanitizeHTMLToDom api, malicious content may still cause data loss!
    • While it is true that meticulously crafted contents may sabotage your data, normal feed providers have no intention to do this (unless they are hacked). So please only subscribe from legitimate websites. I choose not to only show plain text for the contents, because that would ruin the reading experience (we need the outgoing links, images, etc.). Thanks to the reviewer for letting me know of the sanitizeHTMLToDom api!
  • Manage is for managing your feeds, where you can mark all as read, purge those that are marked as deleted, and purge all (i.e. to permanently remove all items of a feed). Be cautious that these actions cannot be undone.
  • All the subscribed sources are listed under their respective folders.
  • The number of unread and total items are displayed after the feed name.
  • To retrieve new items, click on the unread/total number.
  • For each displayed item, you can click Jot to write a short note to it (to be saved in the .md file or the snippet file), Read to mark it as read, Save to create a standalone .md note for it, Snippet to append the content to a snippets file, Math to render the LaTeX equations, GPT to ask ChatGPT to summarize the content, and Delete to mark it as deleted. All these must be done manually (instead of automatically by the plugin itself). Click on the title to show the item content (if Title only is set), or to open the link (if Show content is set).
    • For summarizing with ChatGPT, you need to provide your own ChatGPT API Key and prompt words in the settings tab.
    • Click on Embed to embed the page pointed to by the item URL as an iframe element.
    • Click on Fetch to fetch the URL and display it. This is for pages that cannot be embedded. Better than nothing.
  • The items are paginated with 20 items per page, which can be changed in the settings.
  • The feeds data are saved as gzipped fragmented .json files. They are intentionally fragmented so that when new items come, old files do not need to be updated (hence do not need to be synced).

Acknowledgment

Known issues:

  • Some item URLs cannot be embedded in the mobile, which I understand as a limitation of the current mobile implementation of Obsidian.