
Hi, I need some help regarding the transliteration filters. I would like to use them to transliterate cyrillic name tags (1251) to latin (1250). I have tried with: name=* {set name='${name|ascii:} '} but i get some weird results. For example tag: name=шума is translated into „SHUMA“ instead of expected „ŠUMA“ Same happens with other cyrillic character too – they are translated but not as I expect. I have set code-page=1250 and 1252 but no luck, always the same result. If I set the code page to 1251 I get cyrillic characters but this is not what i want. Obviously there is some transliteration taking place but in the way I expect, maybe I am doing something wrong? Are there any other settings beside code-page? I see there is a folder named ascii in /examples/chars/ with some .trans file, should I put there something else that determinates the transliteration or is this just an output from internal transliteration engine? Should I set country-abbr or country-name? How can I alter the transliteration rules? Thanks in advance for your answer

Hi Aleksandar & Dimitar The files that control the transliteration are in mkgmap.jar: chars/latin1/row*.trans and chars/ascii/row*.trans and these are gathered from {mkgmap}/resources/chars/... in the build process. If you are using latin1/code-page=1252, it looks up the unicode character in latin1 first and, if not defined there, gets the simple ascii transliteration. If not using another character set, it just uses the ascii transliteration. Regards Ticker
participants (2)
-
Aleksandar Lucic
-
Ticker Berkin