Mixed Script Detector
Some characters in a listing are not what they look like. A full-width comma
， and an ordinary one , are different code points that render almost
identically — and a Cyrillic а inside an English word is identical, full stop.
Neither is visible to proofreading, and neither produces an error anywhere.
🔒 Your data never leaves your browser.
Two problems that look like one
Full-width punctuation is what an East Asian input method produces by default. It is correct in Chinese or Japanese typesetting and wrong in an English listing, where a shopper searching for your product types an ordinary comma and never matches yours. This is a mechanical substitution, so this tool performs it:
， → , · 、 → , · 。 → . · ； → ; · ： → : · ？ → ? · ！ → ! · （ → ( · ） → ) · ［ → [
Lookalike letters are a different matter. Cyrillic а е о р с and
Greek ο are visually identical to Latin letters in almost every font. One inside a
word makes it a different string: search will not find it, deduplication will not merge it, and
spellcheck will not complain. There is no way to see this by looking, which is exactly why a
tool has to do it.
Why leftover text is not "fixed" for you
When whole words in another script turn up — a product name, a unit, half a sentence — this tool reports them and stops. Deciding whether to translate them, transliterate them or delete them requires knowing what they say and who the listing is for. Substituting something would be guessing, and a guess in a product title is worse than a flag.
Every check this tool runs (5 rules)
SCR-P01 Full-width punctuation
Characters such as ，。、；：？！（）【】 look like ordinary punctuation at a glance but are different code points. A listing written in English that carries them will not match a shopper’s search, and some feed processors reject the row outright.
Fix: Replace with the ASCII equivalent; the corrected text is offered below.
Source: Derived from the Unicode script property and the CJK compatibility punctuation blocks. The mapping from full-width punctuation to its ASCII counterpart is the standard compatibility decomposition, and our tests check it round-trips character by character.
SCR-S01 Characters from another script
The text mixes writing systems. In a listing aimed at one market this is nearly always leftover source text — a product name, a unit, a stray clause — rather than something intended.
Fix: Translate or remove the leftover text. Which one depends on what it says, so nothing is changed for you here.
Source: Derived from the Unicode script property and the CJK compatibility punctuation blocks. The mapping from full-width punctuation to its ASCII counterpart is the standard compatibility decomposition, and our tests check it round-trips character by character.
SCR-H01 Mixed scripts inside one word
A single word contains characters from two scripts. Cyrillic а, е, о, р, с and Greek ο are visually identical to their Latin counterparts in almost every font, so this is invisible to proofreading — but the word is no longer the word. Search will not match it, deduplication will not catch it, and spellcheck stays silent.
Fix: Retype the affected word rather than editing it; the substitution is impossible to see.
Source: Derived from the Unicode script property and the CJK compatibility punctuation blocks. The mapping from full-width punctuation to its ASCII counterpart is the standard compatibility decomposition, and our tests check it round-trips character by character.
SCR-W01 Full-width Latin letters or digits
Ａ, Ｂ, １, ２ are full-width forms used in CJK typesetting. They are letters and digits to a human and something else entirely to a parser — a full-width digit will not parse as a number.
Fix: Convert to the ordinary half-width forms.
Source: Derived from the Unicode script property and the CJK compatibility punctuation blocks. The mapping from full-width punctuation to its ASCII counterpart is the standard compatibility decomposition, and our tests check it round-trips character by character.
SCR-N01 Single script
The text uses one writing system plus ordinary shared punctuation.
Source: Derived from the Unicode script property and the CJK compatibility punctuation blocks. The mapping from full-width punctuation to its ASCII counterpart is the standard compatibility decomposition, and our tests check it round-trips character by character.
Characters with no faithful ASCII counterpart are reported rather than substituted. The katakana middle dot ・ is the clearest example: replacing it with U+00B7 would swap one non-ASCII character for another and fix nothing.