Text Case Converter

What Title Case rules do you follow — AP, Chicago, or something else?

We default to AP Stylebook rules: capitalize all words of four or more letters, plus the first and last word regardless of length. Short conjunctions, articles, and prepositions ('a', 'an', 'the', 'and', 'or', 'of', 'in', 'on', 'at', 'to', 'for', 'but') stay lowercase unless they are the first or last word. This is the convention used by the Associated Press, most newspapers, and the majority of web publications. If your style guide requires Chicago Manual rules (where all prepositions stay lowercase regardless of length), you will need to manually lowercase words like 'about', 'above', and 'below' after the automated pass.

Why did converting 'getHTTPResponse' to snake_case produce 'get_http_response' instead of 'get_h_t_t_p_response'?

Because the tokenizer recognizes sequences of uppercase letters as a single acronym token when followed by a lowercase letter. 'getHTTPResponse' splits as ['get', 'HTTP', 'Response'] and reassembles cleanly in any target case. Without this heuristic, round-tripping between camelCase and snake_case in a real codebase becomes impossible — 'getHTTPResponse' would turn into 'get_h_t_t_p_response' and back into 'getHTTPResponse' as 'getHttpResponse', breaking imports. The rule matches how lodash.snakeCase, Google's Closure library, and most Ruby ActiveSupport implementations handle this edge case.

Does this handle Unicode characters like Turkish dotted-i or German ß correctly?

Yes. We use String.prototype.toLocaleLowerCase and toLocaleUpperCase rather than the deprecated ASCII-only versions, so Turkish lowercase 'i' (U+0069) correctly uppercases to 'İ' (U+0130) rather than plain 'I', and German lowercase 'ß' (U+00DF) correctly uppercases to 'SS' in most locales or 'ẞ' (U+1E9E) in newer ones. For text-heavy tasks in non-English languages, always test a small sample first because locale-specific rules vary — Azerbaijani also uses dotted-i, Lithuanian has specific rules for 'i' combining with diacritics, and these edge cases surface only in the specific character combinations that trigger them.

Can I convert a 10,000-line file without crashing my browser?

Yes for most cases. The conversion is a single pass over the string with regex operations that run at roughly 100 MB per second on modern CPUs, so a 10,000-line text file (typically 500 KB to 2 MB) converts in under 50 milliseconds. The bottleneck is usually the DOM update when the result is injected back into the output textarea — large outputs can cause a brief paint pause. For files over 10 MB, prefer a CLI tool (sed, awk, or a Node script using lodash.snakeCase) which operates without DOM overhead and can stream the result to disk.

Will proper nouns and acronyms be preserved when converting from lowercase to Title Case?

No. The algorithm has no knowledge of which words are proper nouns or acronyms — 'nasa' lowercased has identical letters to any other four-letter word, so it converts to 'Nasa' in Title Case rather than staying 'NASA'. If you need to preserve specific capitalization patterns (brand names, acronyms, chemical formulas), either pre-process your text to mark them with placeholder tokens and substitute back after conversion, or do a manual pass on the output looking for the specific strings you know should stay capitalized. This is a structural limitation of context-free case conversion, not something any tool in this category solves automatically without ML.