Not sure if this still works in Svelte 5 (maybe?). It seems that Svelte indiscriminately invalidates the UI if there's an AssignmentExpression or UpdateExpression somewhere in the component. Honestly maybe this works even if `count` is a POJO?

JSのfor-inとfor-ofの違いとして, for-inは右にExpressionを許容する一方for-ofはAssignmentExpressionしか許容しないので, for-in 1 1, 2 2は通るけどfor-of 1 1, 2 2はSyntaxErrorが出るのはこの前知ったんだけど, いまだになぜそうなってるのかがわかっていない

PSA: What to do if you suddenly see Property body[42] of BlockStatement expected node to be of a type ["Statement"] but instead got "AssignmentExpression"

AssignmentExpression があるので、plus は両辺にカッコを付けないと写像になってなさそう zenn.dev/uhyo/articles/javas…

Initializer in VariableDeclaration is exactly the AssignmentExpression, but not the basic Expression, since the basic expression derives sequence, which cannot be directly assigned let x = 1; // works let y = z = 2; // works let q = 1, 2; // doesn't work let g = (1, 2); // works

#python #AssignmentExpression

どうでもいいネタもう一つ。JavaScriptの ?: の真ん中は右端と同じ AssignmentExpression。 CやC ではexpression。なのでC/C では cond ? a, b : c // bまたはcになる と書けるが、JavaScriptではエラー。コンマ演算子はあるので (a, b) とカッコで囲めばCと同じ意味に。

The tree for var a = 2; might start with a top-level node called VariableDeclaration, with a child node called Identifier (whose value is a), and another child called AssignmentExpression which itself has a child called NumericLiteral (whose value is 2). #100DaysOfCode

👍🏼 Another way to think about it: spread is not defined as an operator in spec, it's called SpreadElement, and is just valid element of ElementList, which itself is "all the things allowed between the [ ]" of ArrayLiteral... and SpreadElement is: ... AssignmentExpression

へー。こういうのはできなかったですね。まぁアノテーションしたいくらいなら、AssignmentExpression使うなって気持ちでも良い気がしますけどね >>> if c: int := b: File "<stdin>", line 1 if c: int := b: ^

ECMAScript 2017 Language Specification によると、関数の引数は左から右に評価すると定義されているようです。 参照 →「12.3.6.1 Runtime Semantics: ArgumentListEvaluation」の「ArgumentList : ArgumentList, AssignmentExpression」の導出規則の説明のところ。 ecma-international.org/ecma-…

This is green field sub-syntax. If you ask the people who want no semi there about the case where (under maintenance, someone else, later) adds [sym]... on next line, what do they say? Q about class fields: `class { x = y = 1 }` -- legal or not? AssignmentExpression init value.

I’m ok with extra parenthesis, I’m more concerned about AssignmentExpression braces. But I opened an issue : github.com/prettier/prettier… We’ll see

@heycam Need () around term after there. Any such 'let-in' must be low-precedence AssignmentExpression. @littlecalculist @__davidflanagan