Serbian Cyrillic to Serbian Latin: change the script while preserving the text
This direction is transliteration, not a new language
Serbian Cyrillic and Serbian Latin write the same language. In this direction, the central task is to preserve the words, grammar, names, and meaning while changing their written representation. Do not rewrite vocabulary into Croatian or Bosnian merely because the target uses Latin letters. A script choice also does not require changing Ekavian forms into Ijekavian forms, or the reverse.
For an ordinary personal message, the translated output should retain the same person, tense, degree of politeness, and factual claims. If a conversion changes a promise into a request or replaces a familiar word with a different national-standard preference, it has gone beyond the requested script change.
One Cyrillic letter can become a Latin digraph
Serbian љ, њ, and џ correspond to Latin lj, nj, and dž. “Љиљана” becomes “Ljiljana.” The initial digraph uses title-case spelling, with the first letter capitalized and the second lowercase. Do not produce an all-capital start unless the source and the name’s intended capitalization require it.
For an all-capital source, preserve that style: “ЊЕГОШ” becomes “NJEGOŠ.” A single Cyrillic letter can therefore produce more than one Latin character without introducing an extra spoken sound. Character counts will not always remain identical, so compare words and meaning rather than assuming equal length proves a successful conversion.
Use actual Serbian Latin letters
“Ђорђе” becomes “Đorđe.” The target uses đ, not an arbitrary English-looking replacement such as dj. Č, ć, š, and ž likewise need their actual letters. A technical form may impose a restricted-character fallback, but that is a separate requirement and should not silently redefine the ordinary Serbian Latin output.
Preserve the official spelling used for an identifier when exact matching matters. An inflected personal name in a sentence can be transliterated in its current form, while a database field may require its official base form. Do not strip grammatical endings indiscriminately during a script conversion.
Leave nonlinguistic strings usable
A Cyrillic paragraph may contain an existing Latin brand, an email address, a URL, a file path, or a booking code. These strings are not necessarily material for transliteration. Preserve them exactly when changing their characters would break copying, navigation, or identification. Quoted passages may also need to retain their original script if the quotation’s form is relevant.
Finally, inspect visually similar characters. A Latin a and Cyrillic а can look alike while being different characters, affecting search and exact comparison. Ordinary converted words should use a coherent target script, while intentionally preserved technical strings can remain exceptions. Review the whole message for omissions and altered punctuation, not only the distinctive Serbian letters.
Language references