{} + [] REPL bug?

[hemanth]

In the REPL, on 12+ version, I am seeing the below.

> {} + []
'[object Object]'

^ This should be 0

• Version: 12+
• Platform: 18.7.0 Darwin

[ljharb]

This also manifests when you do console.log({} + []) which prints '[object Object]' instead of 0, in many many node versions.

The direct repl behavior seems limited to 12+.

[ljharb]

I've been informed that in fact this is correct; {} + [] in a statement context is "a block" and then +[], which is 0, but in an expression context (or as ({}) + [] is '[object Object]'.

[hemanth]

Also:

> [{}+{}, {}+[]]
[ '[object Object][object Object]', '[object Object]' ]

[devsnek]

This is not a bug. String({}) is '[object Object]' and String([]) is ''.

[hemanth]

Nods, [{} + []] will be '[object Object]' but {} + [] in REPL should be 0?

[devsnek]

REPL is an expression context, the input is actually ({} + []).

[BridgeAR]

We handle REPL input as expression context as @devsnek pointed out. See

node/lib/repl.js

Lines 355 to 358 in 651088c

	// It's confusing for `{ a : 1 }` to be interpreted as a block
	// statement rather than an object literal. So, we first try
	// to wrap it in parentheses, so that it will be interpreted as
	// an expression. Note that if the above condition changes,

This also applies to e.g., { b: 1 }.b.

[hemanth]

As @ljharb highlighted, on node <v12 {} + [] evaluates to 0 on the REPL [devtools might a different story]

[addaleax]

As @devsnek and @BridgeAR pointed out, this is not a bug, but intentional – the REPL prefers to evaluate as expressions before evaluating as statements. So the current one is the expected behavior, and the pre-v12.x one is considered the buggy one.

I’ll close this as I don’t think there’s anything we can do here.

[ljharb]

hmm - i don't see this REPL change listed in the blog post: https://nodejs.org/en/blog/release/v12.0.0/

Clearly it's the right fix to make, and I assume its omission is an oversight - could someone link me to the PR that changed it?

[hemanth]

Was there a PR for this? @ljharb :D