feat: fully configurable plz menu

[0x61nas]

Description

The goal is to provide full power to our users and make them able to configure and create their custom plz menus easily

in order to improve the project and take it to the next level, this PR introduces a bunch of breaking changes to both the configuration system and the command-line interface.

• Change the plz menu system to read the menu entries from the config file
  • Each entry contains its display string that will displayed in the menu and its operation/args
  • The operation is just a hash map that contains the operation key and its value and the rest of the arguments
  • Each operation has a specific set of pre-defined arguments that can be used to configure the operation handler and change its default behavior.
  • The operation handler will ignore any argument that doesn't know about
  • However, all the argument values will be passed to the template engine to be parsed before they are passed to the operation handler, which means that if there's any template error in any of the arguments the program will exit even if the argument that have the error doesn't needed by the operation handler
  • The currently available operations is:
    • fetch | url: fetch any page from the web(mainly) by HTTP
    • run | command: run command/s
    • file: just display the contents of any local file in the pager
• Remove the following CLI options:
  • --cheat-sh-url
  • --eg-url
  • --man-cmd
  • --cheat-url
• Introduce the following CLI options:
  • -s | --selected-position: used to set the default selected position, the posable values are:
    • start: The first item in the menu
    • center: The middle item in the menu (in case of an even number of items, the first item in the second half)
    • end: The last item in the menu

Motivation and Context

The current implementation for the plz menu is nearly limited (there are some configuration options but not many). and we need to give the user full control. This change addresses this demand by introducing new features, refining existing systems, and optimizing the menu's behavior.

• Close make the plz menu configurable via the config file #96

How Has This Been Tested?

• Just by cargo test --lib
• And try the tool with different possible configurations and ensure that they work as expected

Screensthot

The config file in this example:

check_version = true
check_help = true
check = [["-v", "-V", "--version", "version"], ["-h", "--help", "help", "-H"]]
man_command = "man"
pager_command = "less -R"

[plz_menu]
selected_pos = "Center"

[[plz_menu.entries]]
display_msg = "Show man page"

[plz_menu.entries.operation]
run = "man {cmd}"

[[plz_menu.entries]]
display_msg = "Show cheat.sh page"

[plz_menu.entries.operation]
user-agent = "fetch"
fetch = "https://cheat.sh/{cmd}{?/{subcommand}}{? {args}}"

[[plz_menu.entries]]
display_msg = "Show eg page"

[plz_menu.entries.operation]
fetch = "https://raw.githubusercontent.com/srsudar/eg/master/eg/examples/{cmd}.md"

[[plz_menu.entries]]
display_msg = "Show cheatsheets page"

[plz_menu.entries.operation]
fetch = "https://raw.githubusercontent.com/cheat/cheatsheets/master/{cmd}"

[[plz_menu.entries]]
display_msg = "Show info page"

[plz_menu.entries.operation]
file = "~/info-pages/{cmd}.md"

[[plz_menu.entries]]
display_msg = "The TLDR page"

[plz_menu.entries.operation]
run = "tldr {cmd}"
use-pager = "true"

[[plz_menu.entries]]
display_msg = "The navi fuzzy search page"

[plz_menu.entries.operation]
run = "navi --query '{cmd}{? {subcommand}}'"

Types of Changes

• Bug fix
• New feature
• Breaking change
• Documentation
• Refactor
• Other

Checklist:

• My code follows the code style of this project.
• I have updated the documentation accordingly.
• I have formatted the code with rustfmt.
• I checked the lints with clippy.
• I have added tests to cover my changes.
• All new and existing tests passed.
• Update the readme file
• Update the config/halp.toml file

[codecov-commenter]

Codecov Report

Patch coverage is 68.53% of modified lines.

❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Files Changed	Coverage
src/error.rs	ø
src/helper/docs/handlers/command.rs	0.00%
src/helper/docs/handlers/file.rs	0.00%
src/helper/docs/mod.rs	0.00%
src/lib.rs	50.00%
src/helper/docs/handlers/fetch.rs	83.12%
src/helper/docs/template.rs	96.37%
src/cli.rs	100.00%
src/config/mod.rs	100.00%
src/config/plz_menu.rs	100.00%
... and 2 more

📢 Thoughts on this report? Let us know!.

[orhun]

Hey, thanks a lot for the PR! I will probably get to this during the weekend.

There seems to be some conflicts, can you check?

[0x61nas]

I will probably get to this during the weekend.

take your time

There seems to be some conflicts, can you check?

fixed 👍🏻

[orhun]

Hey, sorry for taking too long but just got some time to review this. I'm actually in an airport currently and trying to wrap my head around these changes 🐻

So first of all, this is quite a large PR which begs the following questions:

1. Can we handle this in a simpler manner?

I just checked the halp.toml and the structure was not really intuitive for me:

[plz_menu]
selected_pos = "Center"

[[plz_menu.entries]]
display_msg = "Show man page"

[plz_menu.entries.operation]
run = "man {cmd}"

[[plz_menu.entries]]
display_msg = "Show cheat.sh page"

[plz_menu.entries.operation]
user-agent = "fetch"
fetch = "https://cheat.sh/{cmd}{?/{subcommand}}{? {args}}"

[[plz_menu.entries]]
display_msg = "Show eg page"

[plz_menu.entries.operation]
fetch = "https://raw.githubusercontent.com/srsudar/eg/master/eg/examples/{cmd}.md"

[[plz_menu.entries]]
display_msg = "Show cheatsheets page"

[plz_menu.entries.operation]
fetch = "https://raw.githubusercontent.com/cheat/cheatsheets/master/{cmd}"

Defining arrays in TOML are a bit weird already and it makes it harder to read when we have granual entries like this, Maybe we should consider merging these under one struct like [[entry]].

Another option is to document these well in README.md and include some examples. But we should definitely look into simplifying this.

2. Can this be handled in separate PRs?

I saw that there are quite a bit of commits and it makes it harder to review everything. Maybe consider squashing some of the commits and/or creating separate PRs for finishing this.

3. Should we release a version before this?

There was some discussion about this but I guess I assumed that this will be a simple change which was lowkey wrong. So, should we go ahead with a release before this breaking change?

---

Overall thanks for creating this! We will definitely need to tweak some parts but in the end I really want to have this feature in halp.

[0x61nas]

1. Can we handle this in a simpler manner?

I also didn't want to be in this form initially, as I mentioned in the related issue (#96) I thought that it probably could be done in a just two-dimensional array.
but as we know the ideas are a thing and implementing these ideas is a different thing, especially when we depend on some third-party code

I found that the 2D array way would convert the code into a complete mess and it'll make it hard to maintain so I dropped that option from my list

I've tried to represent the operation with an enum, we've got a limited number of operations and we know that every operation takes a limited amount of parameters. so this makes perfect sense to use an enum for, and obviously, we won't use the enums in Rust without taking its full power, and this will make our life a bit easier when we need to read these values from the other places in our code so we won't need to check every time we need them. but the short long story the toml deserializer wasn't quite happy with that so I had to take another approach

I just checked the halp.toml and the structure was not really intuitive for me:

You're not alone :)

Maybe we should consider merging these under one struct like [[entry]].

I didn't get what you mean, can you provide some more information?

Another option is to document these well in README.md and include some examples. But we should definitely look into simplifying this.

2. Can this be handled in separate PRs?

I have planned to do this already but I was waiting for this PR to get merged or at least get reviewed, so I can consider the changes when I write.

I saw that there are quite a bit of commits and it makes it harder to review everything.

Sorry about that, am still learning

Maybe consider squashing some of the commits

I'll do that

3. Should we release a version before this?

I guess this's the right move for now, but consider updating the halp plz --help command output in the README.md cause I forgot to do that in #56

https://github.com/orhun/halp/blob/main/README.md?plain=1#L236-L243

Options:
  -m, --man-cmd <MAN_CMD>   Sets the manual page command to run
      --cheat-sh-url <URL>  Use a custom URL for cheat.sh [env: CHEAT_SH_URL=]
+   --eg-url <URL>  Use a custom provider URL for eg pages. [env: EG_PAGES_URL=]
+   --cheat-url <URL>  Use a custom provider URL for cheat sheets. [env: ECHEATSHEETS_URL=]
  -p, --pager <PAGER>       Sets the pager to use
      --no-pager            Disables the pager
  -h, --help                Print help

[orhun]

💀

[0x61nas]

died pr, I was too lazy to work on the suggestions, sorry about that. but I'll try to fix it sometime soon, I hope.

I was using it on my machine all this time btw. and I didn't open the config file, I didn't need to do that beside the initial time 7 months ago. so you can see why I did get lazy.

[0x61nas]

But if someone else can fix the (de)serialization mess, I'll not say no. Am not that TOML enthusiast anyway.

[orhun]

Yeah, I need to read through everything again. I don't quite remember what was the status with this.