Tested behavior
Which whitespace characters are removed?
This tool targets horizontal whitespace while deliberately leaving stored line endings in place. That distinction matters when text comes from PDFs, spreadsheets, web pages, or publishing systems that use visually similar but technically different characters. For cells copied from Excel or a CSV file, the CSV whitespace cleaner cleans only the columns you select and keeps the table structure.
| Character group | Examples | Multiple Spaces mode | All Spaces mode |
|---|---|---|---|
| Regular spacing | Space U+0020, Tab U+0009 | Collapse each run to one space | Remove |
| Non-breaking spacing | NBSP U+00A0, figure space U+2007, narrow NBSP U+202F | Convert each run to one regular space | Remove |
| Other supported spacing | U+1680, U+2000–U+200A, U+205F, U+3000, and U+FEFF (BOM) | Convert each run to one regular space | Remove |
| Text-box line endings | LF U+000A (pasted CRLF/CR becomes LF in the browser) | Preserve | Preserve |
| Structural controls | VT U+000B, FF U+000C, NEL U+0085, LS U+2028, PS U+2029 | Preserve | Preserve |
Converting a no-break character to an ordinary space is a deliberate choice rather than a deletion: normal wrapping returns while the words stay apart. If you are not sure why one gap refuses to break, the non-breaking space guide covers the nbsp meaning, where the character comes from, and when to keep it instead of removing it. When those characters are the only problem, remove non-breaking spaces with the dedicated tool and leave tabs and ordinary spaces untouched.
Example: copied product text
Input: Blue[NBSP][NBSP]shirt[TAB]Large
Multiple Spaces: Blue shirt Large
All Spaces: BlueshirtLargeThe audit above counts the hidden character types before the transformation, so the result is explainable instead of being a blind replacement.
Important limits
- ZWSP U+200B, ZWNJ U+200C, ZWJ U+200D and WORD JOINER U+2060 are preserved. U+FEFF is handled as spacing here: Multiple Spaces converts it; All Spaces removes it.
- The browser text box normalizes pasted CRLF and CR to LF before processing. Both modes keep those LF breaks, but do not preserve the original file's line-ending encoding.
- VT, FF, NEL, LS and PS are also preserved. The line counter counts LF-separated text-box rows; these other controls do not increase it.
- All Spaces mode can join words and change meaning; review the output before publishing it.
- Deleting a space inside a URL is not the same as encoding it; see removing spaces from a link before cleaning an address.
- For tab-stop alignment or paragraph-aware line joining, use the dedicated tools below.