Feature list

Encodings

Feature	Details
Encodings	Big5, GBK, UTF-8, UTF-16 LE, UTF-16 BE, HZ, Shift_JIS, EUC-JP, ISO-2022-JP
Detection	BOM → UTF-16 without BOM → binary files → ISO-2022-JP / HZ / ASCII → UTF-8 → scoring Big5, GBK and Japanese; see Encodings and detection
Missing characters	A substitute first (发 → 發 when writing Big5), then &#NNNN; in web pages and ? elsewhere
Web pages	Rewrites <meta charset> and the XML encoding declaration
Entities	Optionally turns &#20013; and &#x4E2D; back into text
UTF-8 BOM	Optional; UTF-16 always gets a BOM

Traditional ⇄ Simplified

Feature	Details
OpenCC engine	OpenCC 1.4.2 dictionaries, converts phrase by phrase (default)
ConvertZ engine	ConvertZ 8.02's character table and fix lists (7,576 Traditional and 2,879 Simplified rules)
Glyph standard	Taiwan, Hong Kong, OpenCC standard
Taiwan vocabulary	软件 ↔ 軟體, 内存 ↔ 記憶體, 鼠标 ↔ 滑鼠…
Custom dictionary	Takes precedence over the built-in dictionaries, works in both directions
Auto	With the script set to Auto, RustZ converts only text that really is the other script

The app

Page	What it does
File conversion	Drop files or folders, detect encodings, preview side by side, convert in bulk, overwrite with backups or save to another folder
Text conversion	Simplified → Traditional and back, live conversion, one-click clipboard conversion, full/half width, mojibake repair with guessing
Rename files	Convert file and folder names, preview first, flag clashes, then apply
Custom dictionary	Add, search, delete and import entries; test conversions as you type
Settings	Theme, interface scale, background mode, default conversion, ignored file types

Also: light and dark themes (or follow the system), 80–125% interface scale, a window that fits the screen, and staying in the menu bar or system tray.

Command line

Feature	Details
ConCmd switches	/i:, /o:, /f:, /q, /b and /cu work as in ConCmd
Extra switches	Engine /e:, glyph standard /v:, vocabulary /p, subfolders /s, dry run /n, overwrite /y, /force, /bom
Input	Files, folders and wildcards, several at once; - reads standard input
Output	Overwrite in place, or write to a file, a folder or a pattern like "out/*.txt"; - writes standard output

All of it is in Command line.

Safe writes

• Output goes to a new, randomly named temporary file first and is then renamed, so an interrupted run never leaves half a file — and symbolic links planted in a downloaded folder are never followed.
• Overwriting can keep a backup as name.bak (or name.1.bak… if that exists; existing backups are never overwritten).
• A file with bytes that cannot be decoded is not overwritten in place without a backup.
• Files whose bytes would not change are not rewritten, so their modification time stays put.
• Files that look binary (NUL bytes and not UTF-16) are skipped.