🐛 Bug: diff@5.0.0 has a performance bug

[jedwards1211]

Bug Report Checklist

• I have read and agree to Mocha's Code of Conduct and Contributing Guidelines
• I have searched for related issues and issues with the faq label, but none matched my issue.
• I have 'smoke tested' the code to be tested by running it outside the real test suite to get a better sense of whether the problem is in the code under test, my usage of Mocha, or Mocha itself.
• I want to provide a PR to resolve this

Expected

Diff performs well even on large inputs

Actual

diff@5.0.0 has catastrophic performance in certain cases (see kpdecker/jsdiff#239).

On my machine, the diff in this example takes ~45 seconds. If I pin to diff@5.2.0, it's almost instantaneous.

It takes some fairly custom settings to trigger this issue in Mocha, but still, you might as well upgrade to diff@5.2.0.

Minimal, Reproducible Example

.mocharc.js

module.exports = {
  spec: ["test.js"],
  "inline-diffs": false,
  reporter: "spec",
  "reporter-option": ["maxDiffSize=525000"],
};

test.js

const { expect } = require("chai");
const { it } = require("mocha");

// make a string of 25k lines of random numbers:
text1 = new Array(25000)
  .fill(0)
  .map((n) =>
    new Array(20)
      .fill(1)
      .map((n) => Math.floor(Math.random() * 10))
      .join("")
  )
  .join("\n");
// grab the first 10,000 characters:
text2 = text1.slice(0, 10000);

it("diff test", () => {
  expect(text1).to.deep.equal(text2);
});

Versions

mocha: 10.4.0
chai: 4.4.1
diff: 5.0.0
node: 20.10.0

Additional Info

No response

[JoshuaKGoldberg]

Fun. The root issue on Mocha's side is that it pins diff to a specific major version. So this is really a subset of #5114. Marking as a duplicate of that one - thanks for filing!