Generate a path-segment candidate
- Paste a title or text and choose Unicode or ASCII-only output.
- Select lowercase language, separator and an optional cluster limit.
- Generate and review the slug and percent-encoded path segment.
Unicode mode first applies canonical NFC normalization, then native locale lowercase. It retains runs beginning with a Unicode letter or number, followed by letters, numbers or marks. This preserves Tamil letters and dependent signs in ordinary text: தமிழ் அழகு becomes தமிழ்-அழகு. Standalone leading marks, emoji, punctuation, spaces, zero-width formatting characters and joiners delimit or disappear between retained runs. This is a deliberate transformation rather than an exact-text encoding.
ASCII mode applies compatibility NFKD decomposition, removes all combining marks, applies selected-locale lowercase and retains a–z/0–9 runs only. Café becomes cafe and circled ① becomes 1. It does not transliterate Tamil or other scripts, expand German ß to ss or Latin Æ to ae; those characters can disappear or divide runs. Tamil-only input can produce an empty ASCII slug. Empty output is reported without inventing a placeholder. Normalization reference.
Length, encoding and collisions
Non-retained spans separate runs with one chosen separator. Leading/trailing separators are omitted. The optional limit counts native grapheme clusters, including separators, then removes any trailing separator from the clipped result. It does not count UTF-8 bytes, percent-encoded characters or display width. Native segmentation is required only when a limit is selected. Output metrics separately report code points, UTF-16 units and UTF-8 bytes.
The percent-encoded value uses encodeURIComponent on the final slug, suitable as a URL path-segment representation. Copy slug and Download slug TXT return the readable slug. The separate URL-encoded controls copy or download its percent-encoded representation. Component encoding reference. The tool does not check uniqueness, reserved route names, destination constraints or SEO outcomes, and does not create a URL or publish a page. Different titles can produce the same slug, especially after normalization, filtering or truncation. Check the destination before use.
Preview and recover changes
Live preview waits 450 ms and pauses during input-method composition. Turn it off for manual generation. An ASCII-mode warning counts letters or numbers outside a–z/0–9 remaining after decomposition, mark removal and lowercase; it does not count every removed symbol or combining mark. Switch to Unicode to retain other scripts. A truncation notice allows removing the chosen limit.
Local TXT import supports one UTF-8 file up to 400 KB, removes an initial BOM and normalizes CRLF/CR to LF. Invalid files preserve your draft/result. Example, import, large paste, Clear and result reuse offer one undo step; typing or changing options discards it. Clear retains settings. Both encoded and readable exports contain the entire result.
Limits and privacy
Native locale/Unicode data can vary; lowercase may add or expand characters after normalization. Processing uses a local worker with ten-second timeout and valid-Unicode input up to 100,000 UTF-16 units; slug output up to 300,000. Cancel terminates work and edits invalidate results. No upload, draft storage, external API or executable HTML. Original TXT preserves the current textarea value, with browser CR/CRLF normalization disclosed. Unpaired surrogates reject for valid UTF-8 export. Clipboard copying is optional; download or select the output if blocked.