Compare two texts

Ignoring whitespace is what makes a diff readable

Diff two versions of a document where someone reflowed the paragraphs and every single line shows as changed. The actual edit — three words in the fourth paragraph — is buried in five hundred lines of noise. This is the normal experience of diffing prose, and it is why most people give up on diff tools for anything but code.

Normalising whitespace before comparison collapses runs of spaces, strips leading and trailing space, and makes reflowed text compare equal to its original. The single genuine change then stands alone. It is the difference between a tool that answers your question and a tool that reprints your document in red.

The algorithm underneath is a longest-common-subsequence match, which finds the smallest set of insertions and deletions that turns one text into the other. It is the same approach Git uses. Its weakness is that a line moved from top to bottom shows as one deletion and one addition rather than a move, because subsequence matching has no concept of relocation.