Text Compare

The Original text is the earlier version; Updated text is the revision. Each upload card also accepts a dropped text file. Supported imports include TXT, CSV, JSON, XML, HTML and Markdown saved as UTF-8. The editor shows both texts before comparison. Swap texts reverses their roles, so additions and removals reverse too. You can compare an empty version with a nonempty version to review an entirely added or removed passage.

Words is the default and is useful for prose. Its comparison units include words, whitespace and punctuation. Lines makes lists and text-based code easier to review. Characters exposes small edits; supporting browsers treat a visible character cluster, such as an emoji sequence, as one unit. Changing the mode can change the similarity percentage and the grouping of changes even when the source texts stay the same.

Both ignore settings start off. Ignore letter case treats differently capitalized matching units as equivalent. Ignore extra whitespace normalizes a run of spaces, tabs or newlines inside a comparison unit to one space. Line boundaries still define units in Lines mode, so changing line breaks can remain visible. These settings change matching, not your stored source text; the complete original spelling and spacing remain in the report.

Removed text uses a red highlight; added text uses green. The previous and next change buttons move between adjacent groups of additions and removals. A replacement may be one group containing both a removal and an addition. The summary counts added and removed words separately from the chosen comparison units. Switching views does not change the comparison or alter either source.

The HTML report contains summary statistics, highlighted source texts and a change register. Open it independently in a browser. The Text report includes the summary, each change and both complete sources. JSON data retains settings, statistics, source texts and comparison segments for further analysis. The print button opens a printable report; choose Save as PDF in the browser print dialog when available. Copy summary copies measurements, not a merged final document.

Each source is limited to 1 MB of UTF-8 text. Large character comparisons and comparisons with many tokens switch to Lines with a visible notice. A very large or substantially different middle section can appear as one broad change. In that case similarity is a lower-bound estimate. Long on-screen previews can be shortened, while downloaded reports retain the complete text. Split the sources into smaller sections when finer matching is needed.