[BUG] Very verbose GPU output

[CIAvash]

Describe the bug The output of GPU entry for AMD 6900 XT is very verbose. It shows:

GPU: Advanced Micro Devices, Inc. [AMD/ATI] Navi 21 [Radeon RX 6800/6800 XT / 6900 XT] (rev c0)

Expected behavior More concise GPU output, for example, neofetch shows this:

GPU: AMD ATI Radeon RX 6800/6800 XT / 6900 XT

Environment

• Version used (4.Y.Z) : 4.13.4
• Method of installation (distribution package, PyPI, Homebrew, sources) : AUR
• Hardware type (laptop, server, Raspberry, hyper-visor) : Desktop
• Python version (3.Y.Z) : 3.10.3
• Operating system and version : Arch Linux (Linux 5.16.16-arch1-1)
• Graphical environment name and version : Sway window manager
• Connectivity (off-line, LAN only, Internet access) : Internet

[HorlogeSkynet]

Hi @CIAvash

It's true that Archey does not assume anything about GPU data format, and simply shows it all (ref). On the other hand, Neofetch performs extended parsing, and specifically for your GPU (ref).

I don't really know what we should/could do about it here, because looking at Neofetch code, this really looks like vendor-related parsing (which I'd rather avoid, because of the maintaining burden).

I'm opened to any idea though, including changing or preferring another data source on GNU/Linux platforms.

Cheers 👋

[CIAvash]

hwinfo shows a shorter output:

hwinfo --short | rg -A 1 'graphics card:' | sd '\s+' ' '

graphics card: ATI Navi 21 [Radeon RX 6800/6800 XT / 6900 XT]

lshw:

lshw -C display 2>/dev/null | rg product

product: Navi 21 [Radeon RX 6800/6800 XT / 6900 XT]

That said, how about adding a field to run custom commands? For example a custom_command field. Or even better allow adding custom entries in which users can run custom commands. What do you think?

[HorlogeSkynet]

Thanks for your command outputs.

That said, how about adding a field to run custom commands? For example a custom_command field. Or even better allow adding custom entries in which users can run custom commands. What do you think?

Definitely, I wanted to (more or less) implement something like this in a rather far future, but I'll try to come up with something sooner then :bow:

[HorlogeSkynet]

Hi @CIAvash, please tell me if #113 implementation draft fits your needs :pray:

[CIAvash]

Hi @HorlogeSkynet, yes, it works perfectly. Thanks!