Script to rename from Cyrillic to English

[ivan]

It would have to be n:n mappings because 2 english characters can combine into one cyrillic one and 2 cyrillic ones can combine into one english one.

There is another problem: since are more characters in russian alphabet than in english one, there isn't always an equivalent letter in english. in fact, some characters only affect the way the word sounds, so individually they don't make any sound.

[admin]

It would have to be n:n mappings because 2 english characters can combine into one cyrillic one and 2 cyrillic ones can combine into one english one.

There is another problem: since are more characters in russian alphabet than in english one, there isn't always an equivalent letter in english. in fact, some characters only affect the way the word sounds, so individually they don't make any sound.

Yep, no problem.

The new scheme (tomorrow) will be:
/;old1/new1;old2/new2;old3/new3; ... etc... (no limits)
1st char= separator between old and new in each mapping
2nd char = separator between mappings

old and new can have any length! You could even use it to transliterate japanese characters to latin syllables, with on click!

I might even call that new command "Transliterate"...

Looking good?

[ivan]

The theory is looking good. I guess I won't know for sure until I try it out. I'm feeling optimistic though.

Another hiccup that I foresee is that letters "s" and "h" (for example) have their Russian individual equivalents. However, when combined ("sh") they can either produce a new letter in Russian alphabet or they can stay together and form a logical part of a word. And that's not the only letter combo that can produce such results. Russian language is very much phonetic so almost each letter is composed of 2 sounds.

How will your transliteration scheme handle that?

[admin]

The theory is looking good. I guess I won't know for sure until I try it out. I'm feeling optimistic though.

Another hiccup that I foresee is that letters "s" and "h" (for example) have their Russian individual equivalents. However, when combined ("sh") they can either produce a new letter in Russian alphabet or they can stay together and form a logical part of a word. And that's not the only letter combo that can produce such results. Russian language is very much phonetic so almost each letter is composed of 2 sounds.

How will your transliteration scheme handle that?

I think this type of problem is not solvable without human intelligence (or something equivalent, which is not visible at the moment). Of course, you could raise the level and work with syllables or even words, but that would mean long long mapping lists...

[ivan]

A possible solution that comes to my mind is to have more than one set of characters associated with certain Latin characters. And, during the rename process, prompt the user which one he/she wants. It gives choice and involves human intelligence...one hopes...

[admin]

A possible solution that comes to my mind is to have more than one set of characters associated with certain Latin characters. And, during the rename process, prompt the user which one he/she wants. It gives choice and involves human intelligence...one hopes...

Yep, that would work. Semi-automatic. But do not expect this from XY -- let's not take this too far...

[ivan]

I guess it would depend on how you plan your transliteration to work. If it's just a simple mapping of characters then I guess not. However, if it will be a complex procedure (aka doing a proper job) then some thought needs to be given

[admin]

I guess it would depend on how you plan your transliteration to work. If it's just a simple mapping of characters then I guess not. However, if it will be a complex procedure (aka doing a proper job) then some thought needs to be given

My plans changed: have to postpone the implementation because other things have to be done first.

Don