#confluence-reader

You can't open Confluence pages using EWW because it requires Javascript. And sure, Confluence gives you the ability to create super-duper dynamic pages. But, and thank heavens and all deities for that, most of the times the pages people create are just text and images.

This package lets you read the "export" view of pages in an Emacs buffer, rendering them via shr (the same built-in library used by EWW), with some minor additions like bookmark support and a few commands for easier navigation.
You can also run a search directly from Emacs, getting a table with the pages that match. There's no context, unlike Confluence's search results, only titles.

#Configuration

(use-package confluence-reader :load-path "/path/to/this/repo"
  :custom
  (confluence-host "thecompany.atlassian.net")
  (confluence-buffer-name-style 'page-title)
  :commands
  (confluence-search confluence-page-by-id confluence-page-from-url))
  

You need to create a token in JIRA and add it to any auth-source enabled location, for example in your .authinfo or .authinfo.gpg file:

machine thecompany.atlassian.net login youruser@thecompany.com password your-token-here

#Customization options

• confluence-host: The host of the Confluence server. Usually has the form "orgname.atlassian.net". This name is also used to lookup credentials using `authsource'."

• confluence-browser-url: URL template to open a page in an external browser. You probably won't need to change this.

• confluence-buffer-name-style: The page buffers are named *Confluence: ?* where ? is affected by this setting. The default page-id makes buffer names shorter. With page-title, they are easier to identify, but they can be very long.

#Usage

This package has three entry points:

• confluence-search: enter some text, get results, hit RET on any of them to display the page. By default it uses standard "text contains" search, invoke with prefix arg to type your own CQL with advanced operators.

• confluence-page-by-id: if you have a page id (you can see them in the URL when on the browser) you can use this to jump directly to that page. It isn't very practical :) but it is used internally by the other commands.

• confluence-page-from-url: just paste the URL from your browser in the minibuffer and this will get the page id from it (using the very scientific method of splitting by "/" chars)

Page view uses confluence-page-mode, where you can see the page exported in glorious plain HTML. Some of the main bindings in this mode are:

• q, the usual "quit window" command
• RET over a link to open it. More about this below.
• b calls bookmark-set, to save the current page in the standard Emacs bookmarking facility.
• TAB to jump through the page's links (add shift to go backwards)
• o to open the page using browse-url-secondary-browser-function (in my case, Firefox)

When rendering a page, this package tries to identify links to other Confluence pages to handle those internally. The internal links have a "➡️" character, while the ones handled by the usual browse-url machinery have "🔗" next to them.
Table of content/anchor links aren't supported, so they get disabled (but still have an underline). You can also access all page headings using imenu.