Smart tokenization
The tokenizer recognizes separators, camel/Pascal humps, acronym runs such as XMLHttp, and optional letter↔digit boundaries before rebuilding the identifier.
Tokens—
Acronyms—
Characters—
Unicode mode—
Recommended—
Warnings—
All developer cases
Language-style audit
Detected tokens
| Line | Input | Output | Tokens | Collision | Audit |
|---|
| # | Token | Normalized | Acronym | Kind | Code points |
|---|
Acronyms are a policy choice
HTTPResponse versus HttpResponse are both used in real codebases. Preserve mode keeps configured initialisms uppercase where the target style allows it; normalize mode treats them as ordinary words.
Unicode versus ASCII-safe
Unicode mode keeps letters from any script. ASCII-safe mode transliterates common Latin, Greek and Cyrillic characters for ecosystems that require conservative identifiers.
Language recommendations
The preset panel gives practical naming-convention guidance, but project-specific linters and style guides remain authoritative.
Collision-safe batch mode
Different source names can normalize to the same identifier. Batch mode flags collisions and can append deterministic numeric suffixes rather than silently overwriting names.
Privacy
All conversion, auditing and CSV export happen locally. No source identifiers are sent to a server.