Invisible Character Finder

Locate selected invisible Unicode and control characters with names and explicit text offsets.

Processing stays in this browser. Input is limited to 100,000 characters.

Reports selected Unicode formatting characters, nonbreaking spaces, tabs and C0/DEL controls. Ordinary spaces, LF and CR are excluded. This is not exhaustive Unicode security analysis.

Inspect hard-to-see characters

Locate selected invisible Unicode and control characters with names and explicit text offsets. Paste text into Input text, then select Process text. This operation has no additional settings. The original input is retained so you can inspect the result before copying or downloading it. Changing input or an option clears the previous output and download.

A reproducible example

Load the example. Three findings appear: U+200B ZERO WIDTH SPACE, U+00A0 NO-BREAK SPACE and U+0009 CHARACTER TABULATION. They have zero-based UTF-16 offsets 1, 3 and 5. The example button replaces both the current input and its settings; use it before entering important work, or keep a source copy elsewhere.

Exact behavior and limits

Reports selected Unicode formatting characters, nonbreaking spaces, tabs and C0/DEL controls. Ordinary spaces, LF and CR are excluded. This is not exhaustive Unicode security analysis. The report has code point, descriptive name, UTF-16 offset, code-point offset, line and column. Offsets begin at zero; lines and columns begin at one. Columns count code points, not visual width or grapheme clusters. An emoji before a finding can increase the UTF-16 offset by two but the code-point offset by one. CRLF counts as one line break. Joiners may be essential for scripts or emoji: a finding is not a recommendation to delete it.

The tab-separated report is generated text; it is not a cleaned copy of the input. The input remains unchanged. The recognized set includes the listed controls plus soft hyphen, selected bidirectional controls, word joiner, zero-width joiners, narrow no-break space and BOM. It excludes many other Unicode characters; a zero-finding result does not certify that text contains no confusing or visually hidden content.

Troubleshooting and reviewing output

An empty input is rejected, but ordinary spaces can produce a valid zero-finding report. Tabs are findings whereas CR and LF only advance report positions. Input is bounded to 100,000 UTF-16 code units. Use the offset convention appropriate to your editor; a visual cursor position can differ from both numeric offsets.

Use Text Compare to inspect edits after you decide which characters are unintended. This finder deliberately reports positions rather than deleting characters automatically, because joiners and bidirectional markers can be meaningful.