feat(filter): --cachedir-ignore option

[SIGSTACKFAULT]

Description

Fixes #948.

adds the --cachedir-ignore option which causes eza to ignore directors with a CACHEDIR.TAG and the correct magic number. see https://bford.info/cachedir/

How Has This Been Tested?

nix flake check passes, but I haven't written formal tests. that's why it's a WIP.

TODO

• Add tests
• Add completions for bash, zsh, fish, nushell
• Add documentation to the man page
• Add your option to the help flag
• Add your option to the README.md

[SIGSTACKFAULT]

oop i thought adding [WIP] made it a draft automagically

[SIGSTACKFAULT]

I added a directory with a CACHEDIR.TAG to itest; should I update all the other tests or put the new directory somewhere else?

[SIGSTACKFAULT]

might change it from --cachedir-ignore to --ignore-cachedir because i was copying --git-ignore but it makes more sense as --ignore-cachedir

[MartinFillon]

Hey, first thing, thanks for the pr, I am going to take a deeper look at it wen I have time. Please make sure to write tests cases and rerun the necessary tools.

[PThorpe92]

Awesome, thanks for the PR!

I like the idea for sure, I'll also check it out later on

[MartinFillon]

As this is a feature that reads data could you make it a feature in cargo toml, make it a default, just to be sure it can be disabled by user not wanting any of that

[SIGSTACKFAULT]

Hey, first thing, thanks for the pr, I am going to take a deeper look at it wen I have time. Please make sure to write tests cases and rerun the necessary tools.

That was one of my questions -- to add a test case i would have to add stuff presumably to the itest directory, but doing that would break all the existing tests. should I update all the existing tests? create a new directory? some third thing?

[SIGSTACKFAULT]

As this is a feature that reads data could you make it a feature in cargo toml, make it a default, just to be sure it can be disabled by user not wanting any of that

as in, a feature that deletes the flag entirely, makes the flag a no-op, makes it not check the magic number, or what? it already only bothers to check for CACHEDIR.TAG if the flag is passed.

[MartinFillon]

as in, a feature that deletes the flag entirely, makes the flag a no-op, makes it not check the magic number, or what

yes

[SIGSTACKFAULT]

which one do you mean by "yes"? i don't understand

[MartinFillon]

All of them, as I already had previous users reaching out on adding features needing dependencies or having small systems, I think that having the option to disable file reading is not a bad idea, if you have thought lmk.

[cafkafk]

I'd also like to add that before I'd consider this ready to merge, the 16 commits should be squashed into more logical units.

[PThorpe92]

Hey finally had a chance to look this over. This looks good man. 👍 Just the one minor fn param.

Seeing as this needs the flag to be enabled, I dont see a reason why it needs to be behind a feature flag.

I do agree with cafk, if we could rebase some of those commits into a couple/few logical conventional commits. This is good to go for me 👍

[MartinFillon]

Seeing as this needs the flag to be enabled, I dont see a reason why it needs to be behind a feature flag.

I will take a look at the code to be sure but I trust you as you read the code, didn't had time for the moment tho.

[SIGSTACKFAULT]

is that squashed enough or do you want more?

[MartinFillon]

Looks good enough

[MartinFillon]

Could you please add at least testing files to the test generator, and if possible please run the powertest binary at the root at the repo, for an example of the way to follow please see #959

[SIGSTACKFAULT]

Yeah, i know it needs tests. i've been nerd-sniped by bevy, sorry >.<

[SIGSTACKFAULT]

just itest fails on main so i think i'm waiting for #959