Text diff

How the comparison works

The two texts are compared as sequences of lines using a longest-common-subsequence algorithm — the same idea behind diff and behind the file comparison in every version control system. It finds the largest set of lines that appear in both, in order, and everything else is an insertion or a deletion.

That is better than comparing line 1 to line 1 and line 2 to line 2, which is what a naive comparison does. Insert one paragraph at the top of a document and a naive comparison says every line changed. This says one line was added.

Where a deletion is immediately followed by an insertion, the two are treated as a rewrite of the same line and the differing words inside them are highlighted. A typo fix shows as a typo fix rather than as two entirely different lines.

The two switches

Ignore capitals compares in lower case. Useful when a document has been through a style pass that changed headings, and you want to see whether anything else moved.

Ignore spacing trims each line and collapses runs of spaces into one. This is the switch that saves you when a document has been through a system that re-wrapped it, or when one copy has tabs and the other has four spaces. Without it, a re-indented file looks entirely rewritten.

Both switches affect the comparison, not the display: what you see is still the original text.

What people use it for

• Contracts. The version they sent back, against the version you sent. This is the most common use and the one where word-level marking earns its keep — a changed "shall" to "may" is a single word in a wall of text.
• Configuration. The file on the machine that works, against the file on the machine that does not.
• Drafts. What the editor actually changed.
• Data. Two exports of the same list, to find the rows that moved.

Limits, and what it will tell you about them

The comparison builds a table of every line on the left against every line on the right. Two thousand lines a side is four million cells, which is fine. Ten thousand a side would be a hundred million and would make a phone unhappy, so past four million cells the tool falls back to a straight line-by-line comparison and says so in the summary rather than freezing the tab.

This compares text, not documents. To compare two PDFs or two Word files, copy the text out of each first — the formatting will not be compared either way.

Nothing is uploaded. Both texts are handled entirely in this tab, which for the contract case is the whole point.