Increase the capacity of "Last Action Shortcut" menu

[kerembaydogan]

Hi thanks for this great plugin.

I need one small improvement.

"Last Action Shortcut" menu is awesome but I think the plugin should at least list two or three last used actions.

In current implementation; if I am using two actions, one after another, I can never use the Last Action Shortcut.

Thanks

Btw: I can write the code and open a pull request if it is ok.

[krasa]

Good idea. Might as well be three.

Sure, you can try.

[gander]

While I don't have Java expertise (it's not my primary language), I recognize the value this feature would bring to the daily workflow with String Manipulation.

I understand implementing promises can be challenging, especially without domain knowledge. Since direct code contribution might not be feasible without proper verification, let me propose two alternative approaches that could achieve similar goals:

Option A: Favorites Submenu
A configurable submenu for frequently-used transformations. Given StringManipulation already has an impressive collection of transformations (50+ actions across case switching, escape/unescape, encoding, sorting, filtering, etc.), having quick access to personal favorites would be invaluable. Users could pin their most-used conversions (e.g., snake_case toggle, JSON escaping, etc.) for instant access.

Option B: Configurable History Size
A settings option to specify how many recent actions appear in the "Last Action Shortcut" menu (e.g., 2-5 items instead of just 1). This would be particularly useful for workflows that alternate between 2-3 transformations repeatedly (like switching between camelCase/snake_case/kebab-case conversions).

I believe Option B might be simpler to implement and would directly address the original issue raised by @kerembaydogan - the current single-item history breaks down when alternating between two actions.

If you're currently swamped with other priorities (no judgment - I completely understand limited bandwidth), and you're open to a PR that leverages AI-assisted development with thorough manual review and testing, I'm willing to take this on. I've learned from past experiences with AI-generated code ("AI slop" won't happen here), so every change would be carefully validated before submission.

The commitment would be mutual: I'd invest time in creating a quality PR, and you'd allocate time for review and feedback. No rush on either end, but wanted to gauge interest before diving in.

What are your thoughts on these approaches?

[krasa]

I already have option B implemented. I was just thinking about making this feature as a freemium, perhaps just only with a nagging notification from time to time, like a WinRAR...

Option A sounds interesting, didn't think of that.

Thank you

[krasa]

Option A: Favorites Submenu

• not sure how pinning would look like from UI perspective

Contributions are welcome, not sure if Ai can do it though