* Revert "Move subclass inheritance to end (#7772)"
This reverts commit f8ab9466d331871a90f458af40b14e8d831e0c29.
* Only use getPrototypeOf if setPrototypeOf is implemented
* Update fixtures
* Helpers updates
* Update fixtures
* Fall back to getPrototypeOf
* Update fixtures
We were using `Object.create` to setup the prototype chain at the start of the class definition, which lead to #7771.
I was a bit worried about a speed hit, but it seems everyone optimizes the two patterns the same way.
https://jsbench.github.io/#f9fca52407643d96458a35763b201215Fixes#7771.
* #7349
- Add "use strict" directive to transformed class bodies
* Update tests (#7349)
* Add test for program with use strict directive (#7349)
* Update tests (#7349)
* Update source mapping (#7349)
* Add test for parent blockStatement with use strict directive (#7349)
* Update tests (#7349)
* Update tests (#7349)
* Update tests (#7349)
* Add test for constructor-only class (#7349)
* Constructor only classes are strict, too
But constructor only that use non-simple parameters must use a strict function wrapper.
* Fix test
* Wrapper not needed if class is already strict
* Revert change to lerna.json
This commit introduces 4 changes:
1) Function declarations are wrapped using function declarations.
This has two advantages:
- We can rely on native hoisting, instead of using _blockHoist
- The function isn't wrapped until it is called. This avoids
problems where `regeneratorRuntime.wrap` was called before
that `babel-polyfill` was imported.
Example:
function fn() {}
// becomes
function fn() { return _fn.apply(this, arguments); }
function _fn() {
_fn = _wrapper(/* Original function ... */);
return _fn.apply(this, arguments);
}
2) Use a single template for both named and anonymous function
expressions. They already had the same behavior, but the one
used for named functions was a bit longer.
3) Use normal functions instead of arrow functions to wrap
function expressions.
4) Generate a name based on the original one for wrapped
functions (e.g. `foo` becomes `_foo` instead of `_ref`).
* Remove whitespace generation and rely on default printing
Changes to printing:
* Add newline after last empty SwitchCase
* Add newlines around block comments if they are non-flow comments or contain newlines
* Fix a few more fixtures
* moved applying arguments inside new Promise handler block
* updated test fixture to reflect change
* corrected the apply to use correct scope and arguments
* added regression test for issue #4943, added async-to-generator/async-default-arguments test
* added regression test for issue #4943
* switched back to using arrow function, since now pointing to v7 release base branch
* simplified async-to-generator regression test for issue #4943, imrproved change to self/arguments refs by using arrow function on returned promise
* updated text fixtures
* removed es2015 preset usage from issue #4943 regression exec test
* added use strict to test fixture
* added use strict to test fixture
* added destructing transform to test options
* removed use strict from exec test
* added parameters & destructing transforms to test
When you write
```
for (const x of l) {
setTimeout(() => x);
}
```
we need to add a closure because the variable is meant to be block-scoped and recreated each time the block runs. We do this.
However, we also add the closure when no loop is present. This isn't necessary, because if no loop is present then each piece of code runs at most once. I changed the transform to only add a closure if a variable is referenced from within a loop.
This change ensures that when hoisting an identifier we do not hoist it up to
the first no shadowed function, but rather up to the bound shadowed function
* Flip default parameter template
YMMV, I saved ~10b on a 2kb library. Not noticeable at the small scale, by why not do it anyway?
I've (unscientifically) found that flipping the default parameter conditional yields better gzip results. I think this is due to the slightly longer string it can now repeatedly match:
```js
// old
var param = arguments.length <= 0 || void 0 === arguments[0] ? null : arguments[0]
--------------------------------------------------------------^
// new
var param = arguments.length > 1 && void 0 !== arguments[1] ? arguments[1] : null
------------------------------------------------------------------------^
```
Though it's entirely likely gzip will also choose up to the index of the arguments if you many default parameters at different indexes.
* Update tests
* Removed unnecessary 'return' statements.
Returning a 'Promise' value in 'promise.then()' makes promise chains.
Used memory and promises are not garbage collected
until finishing 'helpers.asyncToGenerator'.
* Update test
* Fix class inheritance in IE9 & IE10 (T3041)
Internet Explorer 9&10 do not support __proto__ at all, don't have
Object.setPrototypeOf(), but have Object.getPrototypeOf().
Because of this setting the prototype is not possible, which makes the
babelHelpers.inherits() function to set __proto__ although not supported.
Afterwards Object.getPrototypeOf() is used, but this one is not
respecting the "custom" property __proto__ that we set.
The solution is to check for __proto__ first and afterwards fallback to
Object.getPrototypeOf().
* Do the same logic in babel-helper-replace-supers
* Fix tests
* Extract creation of prototype nodes to small helper
