`esprima` AST Tree：如何轻松检测和添加函数括号？

Question

TL;DR: i want to do same thing as there https://github.com/nolanlawson/optimize-js but with esprima when i traverse through AST tree with estraverse .

ESPrima gives same output nodes for following code:

!function (){}()

and

!(function (){})()

http://esprima.org/demo/parse.html?code=!function%20()%7B%7D()%0A%0A!(function%20()%7B%7D)()

For example - i will traverse through AST tree. On Function Expression ExpressionStatement node i want to check - if that node doesn't have parens around function - i want to add it.

So, how i can detect function parens, how i can add them? I look at tokens, but i have no idea how i can associate flat tokens object with object with specified AST node.

Answer 1

似乎它的任务不是针对 esprima，而是针对 escodegen https://github.com/estools/escodegen/issues/315