What invisible characters does it fix?

It converts non-breaking spaces (U+00A0), narrow and thin spaces, ideographic spaces, and zero-width spaces and joiners into ordinary ASCII spaces or removes them. These often sneak in from PDFs, word processors, and pasted web content.

What is a zero-width space?

A zero-width space (U+200B) and its relatives take up no visible room but still count as characters. They break search, sorting, and equality checks, so this tool removes them entirely rather than turning them into visible spaces.

Does it change line breaks?

It preserves line breaks but can collapse runs of three or more blank lines down to one, and trims trailing spaces at the end of each line. Spaces and tabs within a line are collapsed to single spaces when that option is on.

Why does normalised text matter?

Inconsistent whitespace causes duplicate-looking entries, failed string matches, and broken layouts. Normalising to plain single ASCII spaces makes text comparable, searchable, and clean for storage or display.

What is the Whitespace Normaliser?

Clean up messy whitespace: collapse repeated spaces, trim ends, convert non-breaking and zero-width spaces to plain ASCII spaces, and tidy blank lines. Runs entirely in your browser. It runs free in your browser on Gera Tools, with nothing uploaded.

Whitespace Normaliser

Name: Whitespace Normaliser
Creator: Gera Tools
License: https://creativecommons.org/licenses/by/4.0/

Pasted text often carries invisible baggage: non-breaking spaces from PDFs, zero-width characters from web pages, double spaces, and trailing whitespace. This tool normalises all of it to clean, predictable ASCII spacing.

How it works

The cleaner applies a sequence of substitutions:

1. Remove zero-width chars: U+200B, U+200C, U+200D, U+FEFF
2. Map exotic spaces to a plain space:
   U+00A0 (NBSP), U+2007, U+202F, U+2009, U+3000 …
3. Replace tabs with a space
4. Collapse runs of spaces to one (optional)
5. Trim leading/trailing whitespace per line (optional)
6. Collapse 3+ blank lines to a single blank line

The result uses only ordinary U+0020 spaces and newlines, so the text becomes consistent and safe to compare or store.

Example and tips

A heading copied from a slide deck that reads Quarterly Report with a hidden non-breaking space and trailing spaces becomes a clean Quarterly Report. If two strings look identical but fail an equality check, run both through here first — an invisible zero-width character is the usual culprit.

Where invisible whitespace comes from

Different sources introduce different kinds of problem whitespace:

PDFs — text extracted from PDFs often uses U+00A0 (non-breaking space) instead of a regular space, or ligature characters that split awkwardly. Copy-paste from Acrobat Reader is a common source.
Word processors — Microsoft Word and Google Docs both insert non-breaking spaces when you use a specific keyboard shortcut or their autocorrect rules. Pasting into a plain-text field carries these across invisibly.
Web pages — HTML non-breaking spaces ( ) survive copy-paste as U+00A0. Some rich-text editors insert zero-width joiners (U+200D) around emoji or between styled spans.
Spreadsheet exports — CSV or TSV content can have leading/trailing spaces on every cell, causing JOIN and VLOOKUP failures that are very hard to diagnose visually.
AI-generated or translated text — some translation pipelines insert U+200B (zero-width space) at potential break points, or preserve the non-breaking spaces of the source document.

When normalised whitespace matters most

Code and configuration

A stray non-breaking space in a YAML file, a .env value, or a JSON string causes parse errors or comparison failures. If a config value “looks right” but the app rejects it, paste it here first.

Databases and search

Two strings that look identical can be stored as different values if one has a trailing space or an NBSP in the middle. SQL equality checks, Elasticsearch term queries, and unique constraints all treat them as distinct — normalising before insert prevents phantom duplicates.

Spreadsheet formulas

Excel =VLOOKUP() and =MATCH() fail silently when one of the compared strings has a leading space. =TRIM() in Excel handles regular spaces only; it leaves U+00A0 intact. Paste the column through this tool to strip everything, then reimport.

Copy editing and publishing

Automated tools that check for double spaces or flag widows/orphans can be confused by non-breaking spaces that are visually indistinguishable from regular ones. Normalising the draft before running the checker avoids false negatives.

Common mistake: TRIM removes the wrong thing

The standard TRIM() function in Excel, SQL, and most programming languages removes only ASCII U+0020 spaces from the start and end. It leaves U+00A0 and zero-width characters completely untouched. If TRIM() appears to have no effect, the whitespace is almost certainly one of these non-ASCII variants — which is exactly what this tool removes.