Skip to content

Latest commit

 

History

History
74 lines (50 loc) · 3.19 KB

README.md

File metadata and controls

74 lines (50 loc) · 3.19 KB
Raw

bunny

Smart bookmarking tool, running custom commands to open urls from a browser url bar

What?

"Smart bookmarking"? What? I had the same reaction. Imagine it as a very fast, and elegant (if you so wish) way of entering urls into your browser. Instead of writing reddit.com/r/programmerhumor, you could be writing rd r programmerhumor and be redirected to your entertainment of choice a lot easier.

This was inspired by the following article by facebook

How to use

  • cargo install bun (this does require rust nightly for now)
  • Run the installed program bun
  • Create a custom search engine for your browser and point it towards the address the server is running on

The server looks for a file that contains all the bookmarks inside the home directory (~/bookmarks.toml)

Writing bookmarks (bookmarks.toml)

The engine is simple, it's all written in a toml format for readability

  • The bookmarks file is made of books, they can have any name you choose, an alias, and a default url:
[twitter]
alias = "tw"
default = "https://twitter.com"
  • Each book in the file has pages. Each page has a name, a prefix, and a url:
[twitter]
alias = "tw"
default = "https://twitter.com"

[twitter.pages]
search = { prefix = "NONE", url = "https://twitter.com/search?q={encoded}" }
profile = { prefix = "@", url = "https://twitter.com/{raw}" }
  • Each url can contain special keys that handle the data you pass to the command.

Keys

Consider the following command: tw rust lang. Here are the keys and what they do with the given data. The prefix gets stripped away and we are left to handle rust lang

  • {default} - will be replaced with the default url of the book => https://twitter.com
  • {encoded} - will url encode the data => rust%20lang
  • {raw} - will pass in the raw data without encoding it => rust lang
  • {0} - will pass in the first segment of the data => rust
  • {1} - will pass in the second segment of the data => lang
    • There are up to 4 total segments at the moment, because it felt like more were just too many. These can be used to create more customised commands, such as, a reddit command:
    [reddit]
alias = "rd"
default = "https://reddit.com"

[reddit.pages]
whatever = { prefix = "NONE", url = "https://reddit.com/{0}/{1} }
    • The above can be used as rd r programmerhumor to go to a subreddit or as rd u programmerhumor to go to a user

Prefixes

The url prefix can be whatever you want it to be, it is used to differentiate between each command. You could have -s when searching for something, or full on search if shortcuts aren't your thing.

You dont even need spaces between the prefix and command.

-sheyooo will be split into -s heyooo if the prefix is defined as -s.

There's also one special prefix, NONE, which means the command will not expect a prefix, and encode the url with all the given data.

Full examples can be seen in the example file