Unicode Inspector

What each card shows

• Character: one grapheme, the unit a reader sees as a single character. Invisible characters are shown as their code point in brackets.
• Title: the official Unicode name of the character, when the inspector has it.
• Code points: every Unicode code point in that character, written as U+ followed by hexadecimal. A character built from several code points says so, for example “1 character made of 2 code points”.
• UTF-16: how many UTF-16 code units the character uses. Characters outside the Basic Multilingual Plane, including most emoji and styled letters, use two.
• Marks: how many combining marks the character includes, such as accents or glitch marks.
• With style: appears when you choose a style to compare. It shows what that character becomes, or “Stays plain” if the style has no version of it.

Why styles skip some letters

Choose a style in Compare with a style and look for “Stays plain”. Common reasons:

• The style has no digits, as with italic and script.
• The character is not a basic Latin letter, such as é written as one precomposed character.
• Unicode never encoded that letter in that style, such as a small capital x.

Some letters do change but come from an older block. Script B, for example, is SCRIPT CAPITAL B at U+212C rather than a character in the U+1D400 range, because it was encoded years earlier. The inspector shows the exact code point, so you can see this directly.

Turn styled text back into plain text

Choose Unstyle to plain text to convert styled letters back to ordinary ones where that can be done reliably, for example to use a styled bio in a document, a search box or a form. Your original text is not changed; the plain version appears in its own box with a copy button.

FontBlaze restores a character only when it is the exact output of one of its own alphabets, which are generated from the Unicode Character Database. Each result is labeled:

• Restored: an exact reverse, such as MATHEMATICAL BOLD CAPITAL H back to H.
• Case uncertain: some styles, such as small capitals and filled squares, use one form for both capital and small letters, so the original case cannot be known. The plain letter is right, but you may need to fix its capitalization.
• Check meaning: superscripts, subscripts and letterlike symbols such as ℝ are also real notation. m² becomes m2, which may not be what you want.
• Not reversed: upside-down text reverses the order and uses some ordinary letters, such as u for n, so it cannot be turned back reliably and is left as it is.
• Not restored: characters from styled-letter blocks that FontBlaze does not produce, such as Greek math letters, are left as they are.

A converted character is the correct plain equivalent of the styled one, but the plain result is not always the text that was first typed. Review anything labeled Case uncertain or Check meaning before you use it.

Emoji, flags, keycaps, accented letters and other scripts are never changed. Two extra steps are off unless you turn them on. One removes underline, strikethrough and glitch marks, which also removes accents typed as separate marks. The other applies Unicode compatibility normalization (NFKC), which converts more characters but also changes things that are not styling, such as fractions and ligatures. Unicode warns that NFKC should not be applied blindly, so the inspector lists its changes before you apply it.

How styled text may be read aloud

In Unstyle mode, the inspector lists the official Unicode names of the styled characters next to the plain-text alternative. Some screen readers may announce styled letters by these names, spell them out or skip them, so a name like this can become a long string of words. This is a preview built from the Unicode Character Database, not a screen-reader test. Behavior depends on the screen reader, its version, the language and the settings. Unicode itself says these mathematical characters should not replace styling, so use plain text wherever everyone needs to understand it.

Practical uses

• Check a username for invisible characters before you submit it.
• Spot unusual characters in a styled name, then compare them with the rules the app publishes for that field. The inspector cannot tell you whether an app will accept them.
• Find out why two strings that look identical do not match, for example a Latin a and a Cyrillic а.
• Learn what makes up an emoji sequence.
• Turn a styled name back into plain text before using it in a document or form.

For counts and limits, use the character counter. To copy a specific blank character, see invisible characters.