Remove `indexOf` usage from tests in favor of `includes`

[refack]

• Version: > v4
• Platform: all
• Subsystem: test

Since v4 entered maintenance, do you think we can eliminate most of the 433 instances of indexOf from /test/*.js?
Note: The prefered alternative is assert.strictEqual(foo.includes(bar), true)

Each hit should be evaluated whether it's an includes surrogate, or a real indexOf use:

test\addons\repl-domain-abort

• test.js (1 usage found)

test\async-hooks

• init-hooks.js (2 usages found)

test\common

• index.js (1 usage found)

test\disabled

• test-sendfd.js (1 usage found)

test\doctool

• test-doctool-html.js (3 usages found)

test\internet

• test-dns-any.js (1 usage found)
• test-dns.js (1 usage found)

test\parallel

• test-buffer-fakes.js (1 usage found)
• test-buffer-write.js (1 usage found)
• test-child-process-exec-cwd.js (1 usage found)
• test-console.js (1 usage found)
• test-dgram-error-message-address.js (1 usage found)
• test-domain-top-level-error-handler-throw.js (2 usages found)
• test-domain-uncaught-exception.js (2 usages found)
• test-domain.js (1 usage found)
• test-http-keepalive-maxsockets.js (1 usage found)
• test-http-methods.js (3 usages found)
• test-http-outgoing-first-chunk-singlebyte-encoding.js (1 usage found)
• test-http-parser.js (13 usages found)
• test-http-write-head.js (1 usage found)
• test-net-server-connections.js (1 usage found)
• test-path-parse-format.js (1 usage found)
• test-process-getgroups.js (1 usage found)
• test-repl-tab-complete.js (11 usages found)
• test-tls-interleave.js (1 usage found)

test\pummel

• test-dtrace-jsstack.js (3 usages found)
• test-regress-G H-814.js (1 usage found)
• test-regress-GH-814_2.js (1 usage found)

and a special treat

• test\parallel\test-buffer-indexof.js (292 usages found)

[lpinca]

I have no strong opinions, +0.

[vsemozhetbyt]

(BTW, how about new ES6+ label for such issues and PRs?)

[vsemozhetbyt]

cc @nodejs/lts @nodejs/testing

[refack]

(BTW, how about new ES6+ label for such issues and PRs?)

Since it's not a problem with ES6, I think more like move-to-ES6+?

[jseijas]

IMHO that should be done. Reasons: readability and [NaN].

[refack]

Maybe a "good first contribution" or "code-and-lean" like #12376

[gibfahn]

Obviously doesn't make sense for the cases where you actually use the index, but otherwise SGTM.

[refack]

@gibfahn see new desc