Is it possible to provide user-specified transliterations?

[pjfarleyiii]

I have a text document encoded in UTF-8 which I wish to convert to CP1252, but with some specific transliterations. For example I would like any UTF-8 character U+2190 (Left arrow symbol) to be transliterated to the CP1252 character "«" (Left double guillemet, X'AB'), and the UTF-8 character U-2192 (Right arrow symbol) to be transliterated to the CP1252 character "»" (Right double guillemet, X'BB').

The goal here is to convert certain UTF-8 characters to SINGLE CHARACTER transliterations in CP1252, where I could specify transliteration choices in a companion configuration file of some kind.

I am OK with UTF-8 characters with no direct translation into CP1252 for which I provide no overriding transliteration specification to be converted to "?" (question mark) or any default transliteration already set up in recode logic.

Any advice or assistance in how to accomplish this task is appreciated.

[rrthomas]

I'm sorry, this is not currently possible. I'm happy to leave this issue open. The best I can think of in the mean time would be to systematically search and replace the characters in question with the equivalent character in CP-1252, using e.g. the tr or rpl utilities.

[pjfarleyiii]

Thanks for the prompt reply. You can close this issue, as I have determined that my needs are better served by keeping the UTF-8 text as-is instead of converting to CP1252. It would be an interesting extension to recode's capabilities though, by providing more fine-grained control for users that need it.

[rrthomas]

I'll keep the issue open, as I think the suggestion's a good one, but I'm glad you found you didn't need it this time.