XYplorer Beta Club

admin wrote: ↑30 Apr 2026 14:07 So forget it, there is no way to automate this since there is no information about accelerator groups anywhere in the data.

But how do this menu action works?
It must have that info somewhere, no?

Windows does it.

OK I found a solution (almost).
You gave me an idea..
I replaced all the Greek characters in the translated text with Latin counterparts, added the accelerators in ITT, and for every line I got the index of the accelerator and added an "&" in the same place at the original translation (for the counterpart Greek letter).
I didn't include the lines that got more than 1 accelerator, because some of them got double assignments e.g. the line 262 has accelerators for these letters ['t', 'i', 's', 'm', 'm', 'r', 'r', 'r', 'r', 't', 't', 'p', 'o', 'i', 'h', 'l', 'i', 'i', 't', 'o', 'f', 'p', 'i', 'e', 'n', 'l', 'a', 'n', 'u', 'e', 'r', 'i', 'g', 's', 'r'].
ITT also missed ~25 entries that I have to add manually.
But overall, Success!

"... in the same place at the original translation (for the counterpart Greek letter)."
1) what if the translation is shorter than that place?
2) (worse) you will get many dupes

No, perhaps I didn't explain it properly..
Lets say we have the line
co.0011=01|O&verwrite|--|Αντικατάσταση
I use this table
αβγδεζηθικλμνξοπρστυφχψωςέύίόάήώ
abgdezhqiklmnjoprstufxywceuioahw
to convert it to
co.0011=01|O&verwrite|--|antikatastash
After ITT adds the accelerators it becomes
co.0011=01|O&verwrite|--|a&ntikatastash
The index of the & is 1 (for the translated text), so I add an & at this index in the original translation and I get this
co.0011=01|O&verwrite|--|Α&ντικατάσταση

I get no dupes because there are not random letters, they have a 1to1 relationship with the Greek characters..

Hm.

1) Your table (abgdezhqiklmnjoprstufxywceuioahw) does not contain "v" which is the accelerator in your example.

2) You just put the accelerator at the same position. All the table mapping is not necessary.

admin wrote: ↑30 Apr 2026 19:27 1) Your table (abgdezhqiklmnjoprstufxywceuioahw) does not contain "v" which is the accelerator in your example.

The ν in "Αντικατάσταση" is not the Latin v. It's a Unicode Greek letter that "sounds" like the Latin n.

2) You just put the accelerator at the same position. All the table mapping is not necessary.

The reason I use the mapping is to let Windows calculate what is the best character (Latin) to assign the accelerator.
When it does, I simply swap it with the Greek counterpart.
Keep in mind that the uniqueness of each character is exactly the same in the Greek characters, as in the Latin ones.

Also, do you have any idea why there are dupes in the line 262 of the .lng file (line 245 in the ITT)?

I meant the Latin "v".

Dupes in 262: That's okay. Sometimes it's unavoidable (e.g when there are more than 26 items). In that case you press the key again to jump to the next accelerator.

admin wrote: ↑30 Apr 2026 20:38 I meant the Latin "v".

Oh.. Now I understood!
You mean the v in the "O&verwrite"
Well, I don't change anything at the original part.
I just change things at the translated part.
So in the example, the original translation was "Αντικατάσταση", which change into "antikatastash", that the ITT automatically added the accelerator and made it "a&ntikatastash", and then I saw were ITT put it (index 1 of the word) and replace it with the original translated word plus the accelerator at the same place ("Α&ντικατάσταση")

Dupes in 262: That's okay. Sometimes it's unavoidable (e.g when there are more than 26 items). In that case you press the key again to jump to the next accelerator.

Hmm.. This is interesting. Maybe I'll do that as well.
I have to change the code a little though, we'll see..

FYI, when ITT does not find the letter in the translation (no "v" in "antikatastash") it defaults to position 1 (2nd letter).

admin wrote: ↑30 Apr 2026 21:13 FYI, when ITT does not find the letter in the translation (no "v" in "antikatastash") it defaults to position 1 (2nd letter).

Nope, this isn't the case. Another example.

The original translation
co.0014=01|Resu&me|--|Συνέχεια
I change it to Latin
co.0014=01|Resu&me|--|sunexeia
ITT adds accelerators
co.0014=01|Resu&me|--|su&nexeia
I change it to the original translation
co.0014=01|Resu&me|--|Συ&νέχεια

Ah yes, there is a little smartness built in to avoid immediate dupes with the previous item.

Well, I added the multiple accelerators too. Can you check if this version is OK?

I don't have a Greek keyboard. You should check it.