Case Converter — UPPER, lower, Title, camelCase Online

NFD (Canonical Decomposition, defined in UAX #15 — Whistler ed., revision 57 tied to Unicode 17.0) splits precomposed characters into a base letter plus combining mark — so 'é' (U+00E9) decomposes to 'e' (U+0065) + combining acute accent (U+0301). The combining mark sits in the Combining Diacritical Marks block (U+0300–U+036F), and a regex strip after normalize removes it cleanly. Emoji like 😀 (U+1F600) and CJK ideographs like 漢 (U+6F22) are atomic — they have no canonical decomposition into base + combining mark, so NFD leaves them as a single code point and the diacritic regex doesn't touch them. The four programmer styles (camelCase, PascalCase, snake_case, kebab-case) then strip everything outside ASCII letters and digits, which is when emoji and CJK get dropped — but that's the alphanumeric filter, not the NFD pass.

The snake_case and kebab-case paths split on lowercase-to-uppercase transitions inside the input before joining with the separator. The regex inserts a boundary wherever a lowercase letter or digit is immediately followed by an uppercase letter, so 'myCamelCase' becomes 'my Camel Case' internally, then lowercases to 'my camel case', then joins with '_'. This is the standard convention shared by most snake_case libraries (Ruby's `underscore`, Python's `inflection`, Lodash's `_.snakeCase`) — it preserves the word boundaries an existing camelCase identifier already encodes, rather than collapsing them. The same logic produces 'parse_html_file' from 'parseHTMLFile' — consecutive caps are treated as a single word boundary, and the entire result is lowercased per the snake_case convention.

No — this implementation capitalizes the first letter of every whitespace-separated word and lowercases the rest, which is the simple algorithmic Title Case. English style guides (AP, Chicago Manual of Style, APA) all add an additional rule that articles ('a', 'an', 'the'), short conjunctions ('and', 'but', 'or'), and short prepositions ('in', 'on', 'of') stay lowercase unless they're the first or last word — so a strict-AP version of 'A Tale of Two Cities' is 'A Tale of Two Cities', not 'A Tale Of Two Cities'. The simple version is correct for headlines in CMS systems, code identifiers, and most casual use; for editorial copy targeting a specific style guide, post-processing the output to lowercase the function-word list closes that gap.

JavaScript's String.prototype.toLowerCase / toUpperCase use the default Unicode case mappings. Under default Unicode, dotted İ (U+0130) lowercases to plain i + combining dot above (U+0069 + U+0307) — the combining mark preserves roundtrip information that the source had a dotted capital — and dotless capital I (U+0049) lowercases to plain Latin i (U+0069). Turkish has a different convention: under the Turkish locale, dotted İ lowercases to plain i (U+0069) and dotless capital I lowercases to Turkish dotless ı (U+0131). To get the Turkish convention, applications call String.prototype.toLocaleLowerCase('tr'), which uses the locale-aware mapping coordinated with ECMA-402's locale resolution. This tool ships the default Unicode mapping because Turkish-specific identifier conventions are rare in global naming use cases; for code or filenames targeting Turkish locales, the locale-aware variant covers the gap. The same caveat applies to Azerbaijani (az), which shares Turkish dotted/dotless I.

The output region for each style is marked aria-live="polite", the W3C WCAG Success Criterion 4.1.3 (Status Messages, introduced in WCAG 2.1, Recommendation 5 June 2018; carried unchanged into WCAG 2.2, Recommendation 5 October 2023) pattern. Polite live regions queue announcements after any speech in progress, so switching styles or editing input announces the result without interrupting the user mid-sentence. Screen readers (NVDA, JAWS, VoiceOver) consume the live region automatically; nothing else is required from the user.