Text Diff Checker

What the comparison shows

The first text goes on the left, the second on the right. Matching lines line up directly across from each other, and changed lines are highlighted. Within a changed line, the words that were edited are marked separately, so you see the edit itself and not just the fact that the line was touched.

Below the result are the counts: how many lines match, how many were removed and added, and what share of the text stayed the same.

Differences you can't see on screen

Two lines can look exactly alike while the comparison still counts them as different. This happens when text comes from different sources, and there are a few common reasons. Each one is named right in the results:

• a Latin letter inside a Russian word. The letters «с», «о», «е», «а», «р», «х» exist in both Cyrillic and Latin and look identical;
• «ё» instead of «е», or the other way around;
• a non-breaking space, a soft hyphen, a zero-width space. The characters are there, but the screen shows nothing;
• different dashes and hyphens, different quotes, an ellipsis as a single character instead of three periods;
• extra spaces, including at the end of a line.

The "show invisible characters" checkbox renders these characters as visible marks right in the text, so the spot that was edited can be seen with your own eyes.

Three ways to compare

By line suits contracts, lists, code and data dumps: a line there is a unit of meaning on its own.

By word works better for running text, such as articles, product descriptions, letters. Line breaks there are placed wherever, and a line-by-line comparison would show the whole paragraph as shifted instead of one edit.

By character is for short strings: SKU codes, numbers, keys, addresses.

Relaxing the comparison

Four checkboxes take out of the comparison what plays no role in a particular task: letter case, extra spaces, the difference between «ё» and «е», empty lines. The verdict then honestly notes that there are no differences specifically with these relaxations applied.

Files and encodings

You can drag a file right into the field. If the text comes out as garbage characters, the file is re-read in Windows-1251, which is how old documents and 1C data exports open.

Texts stay in the browser

Nothing goes to the server: both the comparison and the file reading run on your side. That is why contracts, drafts and work documents can be pasted here.