fix: better align access keys with the guidelines

[smileyhead]

Summary

Resolves #678.

Based on Microsoft's guidelines, I have…

• added access keys (mnemonics) to all menu items and some buttons which I found were missing them
• moved the position of existing mnemonics where needed
• created new strings where a mnemonic was needed for only one instance of that string
• created new strings specifically in the New Tag window to replace the generic ‘+’ buttons with something more descriptive that allows for mnemonics
• created strings and added labels to the main window drop-downs to allow for mnemonics
• created links (buddies) between labels and corresponding UI elements to allow for mnemonics
• modified the File -> Open Recent list to prefix numbers in front of the directory paths, which allow for mnemonics
• deleted unneeded _alt strings, as some of them should always have mnemonics and others should never have them

Unresolved issues:

• I would have liked to create a mnemonic for the ‘Color’ label in the New Tag window that activates the button below it. The buddy has been defined, but it does not work and I do not know why.
• The search box cannot be added mnemonics to directly and I am unsure on how to proceed (see this comment for more info)

Tasks Completed

• Platforms Tested:
  • Windows x86
  • Windows ARM
  • macOS x86
  • macOS ARM
  • Linux x86
  • Linux ARM
• Tested For:
  • Basic functionality
  • PyInstaller executable

[smileyhead]

While looking into adding mnemonics to the search bar in the main window, I have discovered that adding them to a QLineEdit element directly is, unfortunately, not supported.

We might consider adding a shortcut key to focus on the text box (Ctrl + F, for instance) and signal this in the placeholder text (‘Search Entries ({shortcut})’), but that is out of scope for this PR. Alternatively, we could add a buddied QLabel in front of the search box, something like ‘Search Entries:’, but I do not wish to do that without prior approval.

[smileyhead]

I apologise; I realised I have made this PR too soon, as there are still more commits I wish to make. I will reopen this when I am done.

[smileyhead]

I have made all the modifications I wanted and updated the PR to reflect that. It is now ready for review.

[CyanVoxel]

I apologize for the inconvenience - a large refactor was recently merged (#844) that has moved several files throughout the codebase. This was done in order to correct some deep issues with the project's layout and was performed now as it seemed to be the best opportunity (as of 3/6/25) to disrupt as few open PRs as possible.

Unfortunately this PR is one that's been disrupted by this change. A rebase targeting main will be required to resolve the merge conflicts generated here. A merge commit is possible, however I do not recommend it as a rebase will be much smoother and much less work. I apologize again for the disruption.

I also noticed that the Ruff workflow was failing beforehand, please let me know if you'd like any help with that.

[smileyhead]

This is the first time I have dealt with this issue, so apologies if it doesn't work the first time. Here goes.

[smileyhead]

I have decided to also address the question of the _alt strings while waiting for a review. The modifications I've made are as follows:

Kept:

• generic.apply_alt: Apply (renamed to generic.apply)
• generic.cancel: Cancel
• generic.done: Done
• generic.delete: Delete
• generic.edit_alt: Edit (renamed to generic.edit)
• generic.overwrite: Overwrite
• generic.rename_alt: Rename (renamed to generic.rename)
• generic.skip_alt: Skip (renamed to generic.skip)

Removed:

• generic.apply: Apply – Reason: Unused, and this should always have the mnemonic Apply by convention.
• generic.cancel_alt: Cancel – Reason: The access key for Cancel is Esc which already works in TagStudio.
• generic.done_alt: Done – Reason: Likewise to Cancel, the access key for Apply is Esc.
• generic.delete_alt: Delete – Reason: Destructive actions should not have mnemonics, so as to avoid accidental activation.
• generic.edit: Edit – Reason: Again, this should have mnemonics.
• generic.overwrite_alt: Overwrite – Reason: Destructive action.
• generic.rename: Rename – Reason: Unused.
• generic.skip: Skip – Reason: Unused.

[smileyhead]

Sorry, is this PR still on the way to getting merged? I totally understand if you're just busy with other things in the project, I'm just not sure how to interpret this silence.

As far as I can tell, the failing check is due to the string changes I've made producing key mismatches across the different language .JSONs. I assume this can be fixed automatically? If not, I could go over all the language files and make the necessary changes manually.

[CyanVoxel]

Sorry, is this PR still on the way to getting merged? I totally understand if you're just busy with other things in the project, I'm just not sure how to interpret this silence.

Yes, sorry about the delay! When I initially pulled this to review I noticed the new mnemonic underlines across the interface of the program, which led me to research what the default behavior of other programs is and how this might be able to be avoided. This led me to create #856 since macOS doesn't support keyboard mnemonics at all and that would at least prevent the UI from being cluttered on that system, but unfortunately I couldn't find anything useful to suggest regarding Qt's handling of mnemonics. It just feels a bit odd to have underlines across most of the UI text when other programs on Windows don't have to deal with this because the underlines only appear (outside of a menubar) when the ALT key is pressed. If there was a straightforward way to do that here then I wouldn't have any complaints, but so far I haven't been able to find any way to do that with Qt. One option I was considering was the ability to make mnemonic underlines a toggle in the settings, which would then use a similar "filtering" method to #856. But since the new settings menu in #859 hasn't been merged yet I also didn't want to tell you to add a setting that would immediately conflict with that PR.

As far as I can tell, the failing check is due to the string changes I've made producing key mismatches across the different language .JSONs. I assume this can be fixed automatically? If not, I could go over all the language files and make the necessary changes manually.

I would appreciate if you could do that, I'm not sure if Weblate would handle doing it automatically very well and it would mean that the workflow checks would be failing on main in the meantime.

[smileyhead]

make mnemonic underlines a toggle in the settings

Being able to always see mnemonics even when not holding Alt is a system setting in Windows. I wonder if it would be possible to make TagStudio only show them as long as Alt it held down while this is disabled? It would go around having to add a manual override on an app level.

I would appreciate if you could do that

Alright, I will do that.

[smileyhead]

Now that I'm looking through the other language files, several of them are already missing lots of strings. I wonder why they weren't triggering a fail before…?

[Computerdores]

Now that I'm looking through the other language files, several of them are already missing lots of strings. I wonder why they weren't triggering a fail before…?

The check validates that the format strings are correct (they use all the keys the english one uses and no others than those) not that all keys have translations. After all, we wouldn't want the checks to fail because a language isn't fully translated yet

[smileyhead]

That should be it, then.

The check validates that the format strings are correct (they use all the keys the english one uses and no others than those) not that all keys have translations.

Ah, I see. That makes sense.