This is just factoring out step "9. Set O.[[Prototype]] to V." of
10.1.2 [[SetPrototypeOf]] into its own method so that we don't have to
use internal_set_prototype_of() for setting an object prototype in all
cases.
The way that transition avoidance (foo_without_transition) was
implemented led to shapes being unshareable and caused shape explosion
instead, precisely what we were trying to avoid.
This patch removes all the attempts to avoid transitioning shapes, and
instead *adds* transitions when changing an object's prototype.
This makes transitions flow naturally, and as a result we end up with
way fewer shape objects in real-world situations.
When we run out of big problems, we can get back to avoiding transitions
as an optimization, but for now, let's avoid ballooning our processes
with a unique shape for every object.
Before this we used an ad-hoc combination of references and 'variables'
stored in a hashmap. This worked in most cases but is not spec like.
Additionally hoisting, dynamically naming functions and scope analysis
was not done properly.
This patch fixes all of that by:
- Implement BindingInitialization for destructuring assignment.
- Implementing a new ScopePusher which tracks the lexical and var
scoped declarations. This hoists functions to the top level if no
lexical declaration name overlaps. Furthermore we do checking of
redeclarations in the ScopePusher now requiring less checks all over
the place.
- Add methods for parsing the directives and statement lists instead
of having that code duplicated in multiple places. This allows
declarations to pushed to the appropriate scope more easily.
- Remove the non spec way of storing 'variables' in
DeclarativeEnvironment and make Reference follow the spec instead of
checking both the bindings and 'variables'.
- Remove all scoping related things from the Interpreter. And instead
use environments as specified by the spec. This also includes fixing
that NativeFunctions did not produce a valid FunctionEnvironment
which could cause issues with callbacks and eval. All
FunctionObjects now have a valid NewFunctionEnvironment
implementation.
- Remove execute_statements from Interpreter and instead use
ASTNode::execute everywhere this simplifies AST.cpp as you no longer
need to worry about which method to call.
- Make ScopeNodes setup their own environment. This uses four
different methods specified by the spec
{Block, Function, Eval, Global}DeclarationInstantiation with the
annexB extensions.
- Implement and use NamedEvaluation where specified.
Additionally there are fixes to things exposed by these changes to eval,
{for, for-in, for-of} loops and assignment.
Finally it also fixes some tests in test-js which where passing before
but not now that we have correct behavior :^).
- Replace the misleading abuse of the m_transitions_enabled flag for the
fast path without lookup with a new m_initialized boolean that's set
either by Heap::allocate() after calling the Object's initialize(), or
by the GlobalObject in its special initialize_global_object(). This
makes it work regardless of the shape's uniqueness.
- When we're adding a new property past the initialization phase,
there's no need to do a second metadata lookup to retrieve the storage
value offset - it's known to always be the shape's property count
minus one. Also, instead of doing manual storage resizing and
assignment via indexing, just use Vector::append().
- When we didn't add a new property but are overwriting an existing one,
the property count and therefore storage value offset doesn't change,
so we don't have to retrieve it either.
As a result, Object::set_shape() is now solely responsible for updating
the m_shape pointer and is not resizing storage anymore, so I moved it
into the header.
This is a tiny difference and only changes anything for primitives in
strict mode. However this is tested in test262 and can be noticed by
overriding toString of primitive values.
This does now require one to wrap an object in a Value to call invoke
but all code using invoke has been migrated.
These were an ad-hoc way to implement special behaviour when reading or
writing to specific object properties. Because these were effectively
replaced by the abillity to override the internal methods of Object,
they are no longer needed.
This removes all usages of the non-standard define_property helper
method and replaces all it's usages with the specification required
alternative or with define_direct_property where appropriate.
Now that the Object rewrite is in place, we have enough tools to
implement the mapped `arguments` propreties according to spec.
The basic mechanism is that the `arguments` object installs a hidden
parameter mapping object that property accesses get filtered through.
This is how accessing numeric properties on `arguments` are proxied
to the named identifier in the function scope.
When `arguments` is instantiated, getters and setters are created
for all the numeric properties on the object that correspond to
function arguments. These getters and setters can be deleted from the
object. This is all pretty intricate, so refer to the spec for details.
Note that the `arguments` object itself is still lazily instantiated
on first access within a function. This is non-conforming, and we'll
have to revisit this once we get around to improving function calls.
To make this happen, this patch implements the SetImmutablePrototype
abstract operation (as a method on Object) and then overrides
[[SetPrototypeOf]] on ObjectPrototype.
This is a huge patch, I know. In hindsight this perhaps could've been
done slightly more incremental, but I started and then fixed everything
until it worked, and here we are. I tried splitting of some completely
unrelated changes into separate commits, however. Anyway.
This is a rewrite of most of Object, and by extension large parts of
Array, Proxy, Reflect, String, TypedArray, and some other things.
What we already had worked fine for about 90% of things, but getting the
last 10% right proved to be increasingly difficult with the current code
that sort of grew organically and is only very loosely based on the
spec - this became especially obvious when we started fixing a large
number of test262 failures.
Key changes include:
- 1:1 matching function names and parameters of all object-related
functions, to avoid ambiguity. Previously we had things like put(),
which the spec doesn't have - as a result it wasn't always clear which
need to be used.
- Better separation between object abstract operations and internal
methods - the former are always the same, the latter can be overridden
(and are therefore virtual). The internal methods (i.e. [[Foo]] in the
spec) are now prefixed with 'internal_' for clarity - again, it was
previously not always clear which AO a certain method represents,
get() could've been both Get and [[Get]] (I don't know which one it
was closer to right now).
Note that some of the old names have been kept until all code relying
on them is updated, but they are now simple wrappers around the
closest matching standard abstract operation.
- Simplifications of the storage layer: functions that write values to
storage are now prefixed with 'storage_' to make their purpose clear,
and as they are not part of the spec they should not contain any steps
specified by it. Much functionality is now covered by the layers above
it and was removed (e.g. handling of accessors, attribute checks).
- PropertyAttributes has been greatly simplified, and is being replaced
by PropertyDescriptor - a concept similar to the current
implementation, but more aligned with the actual spec. See the commit
message of the previous commit where it was introduced for details.
- As a bonus, and since I had to look at the spec a whole lot anyway, I
introduced more inline comments with the exact steps from the spec -
this makes it super easy to verify correctness.
- East-const all the things.
As a result of all of this, things are much more correct but a bit
slower now. Retaining speed wasn't a consideration at all, I have done
no profiling of the new code - there might be low hanging fruits, which
we can then harvest separately.
Special thanks to Idan for helping me with this by tracking down bugs,
updating everything outside of LibJS to work with these changes (LibWeb,
Spreadsheet, HackStudio), as well as providing countless patches to fix
regressions I introduced - there still are very few (we got it down to
5), but we also get many new passing test262 tests in return. :^)
Co-authored-by: Idan Horowitz <idan.horowitz@gmail.com>
Specifically, this now explicitly takes the length, adds missing
exceptions checks to calls with user-supplied lengths, takes and uses
the prototype argument, and fixes some spec non-conformance in
ArrayConstructor and its native functions around the use of ArrayCreate
If we define a property with just a setter/getter (not both) we must:
- take the previous getter/setter if defined on the actual object
- overwrite the other to nullptr if it is from a prototype
In get_own_properties:
Entries which are deleted while iterating need to be skipped
In PropertyDescriptor::from_dictionary
If the getter/setter is undefined it should still mark it as present
The specification defines that we should only change attributes that
exist in the incoming descriptor, but since we currently just overwrite
the existing descriptor with the new one, we can just set the missing
attributes to the existing values manually.
This commit expands on 5eef07d232 by
automatically trying to coerce Type::String PropertyNames into numbers
when a caller checks if the PropertyName is_number/is_string.
This has several benefits:
- We no longer have to duplicate the number coercion code to every
function that accepts a PropertyNumber. (Or more likely, forget to.)
- This keeps the lazy nature of only doing the coercion when and if
there is a semantic difference to the different PropertyName types,
which means this shouldnt cause any performance drop.
- Since this coercion changes the state of the PropertyName itself the
result is essentially cached and can speed up any repeat uses of the
same PropertyName instance.
This now matches the spec's OrdinaryObjectCreate() across the board:
instead of implicitly setting the created object's prototype to
%Object.prototype% and then in many cases setting it to a nullptr right
away, it now has an 'Object* prototype' parameter with _no default
value_. This makes the code easier to compare with the spec, very clear
in terms of what prototype is being used as well as avoiding unnecessary
shape transitions.
Also fixes a couple of cases were we weren't setting the correct
prototype.
There's no reason to assume that the object would not be empty (as in
having own properties), so let's follow our existing pattern of
Type::create(...) and simply call it 'create'.
We were doing a *lot* of string-to-int conversion while creating a new
global object. This happened because Object::put() would try to convert
the property name (string) to an integer to see if it refers to an
indexed property.
Sidestep this issue by using PropertyName for the CommonPropertyNames
struct on VM (vm.names.foo), and giving PropertyName a flag that tells
us whether it's a string that *may be* a number.
All CommonPropertyNames are set up so they are known to not be numbers.
As mentioned on Discord earlier, we'll add these to all new functions
going forward - this is the backfill. Reasons:
- It makes you look at the spec, implementing based on MDN or V8
behavior is a no-go
- It makes finding the various functions that are non-compliant easier,
in the future everything should either have such a comment or, if it's
not from the spec at all, a comment explaining why that is the case
- It makes it easier to check whether a certain abstract operation is
implemented in LibJS, not all of them use the same name as the spec.
E.g. RejectPromise() is Promise::reject()
- It makes it easier to reason about vm.arguments(), e.g. when the
function has a rest parameter
- It makes it easier to see whether a certain function is from a
proposal or Annex B
Also:
- Add arguments to all functions and abstract operations that already
had a comment
- Fix some outdated section numbers
- Replace some ecma-international.org URLs with tc39.es
This is very similar to Object::define_native_property, but here the
native functions are exported as standalone JS getter and setter
functions, instead of being transparently called by interactions with
the property.
This would return an empty value once it hits an exception check
otherwise. Considering that this mostly is used in situations where we
already *do* have an exception (traceback printing, for example), let's
make this easier for ourselves to use.
SPDX License Identifiers are a more compact / standardized
way of representing file license information.
See: https://spdx.dev/resources/use/#identifiers
This was done with the `ambr` search and replace tool.
ambr --no-parent-ignore --key-from-file --rep-from-file key.txt rep.txt *
This was failing to take two things into account:
- When constructing a PropertyName from a value, it won't automatically
convert to Type::Number for something like string "0", even though
that's how things work internally, since indexed properties are stored
separately. This will be improved in a future patch, it's a footgun
and should happen automatically.
- Those can't be looked up on the shape, we have to go through the
indexed properties instead.
Additionally it now operates on the shape or indexed properties directly
as define_property() was overly strict and would throw if a property was
already non-configurable.
Fixes#6469.
Similar to Value::to_string_without_side_effects() this is mostly a
regular object property lookup, but with the guarantee that it will be
side-effect free, i.e. no accessors or native property functions will
be called. This is needed when we want to access user-controlled object
properties for debug logging, for example. The specific use case will be
error objects which will soon no longer have internal name/message
properties, so we need to guarantee that printing an error, which may
already be the result of an exception, won't blow up in our face :^)
This fixes two cases of indexed access (array holes, out-of-bounds
string object access) where we would not follow the prototype chain and
incorrectly return undefined:
// Should be "a", returned undefined
Object.setPrototypeOf([,], ["a"])[0]
// Should be "a", returned undefined
Object.setPrototypeOf(new String(""), new String("a"))[0]
The actual fix is simple, instead of returning early if the requested
index is past the string's length or within the indexed properties size
but has no value, we just continue the prototype chain traversal and get
correct behaviour from that.
This replaces the current 'function plus boolean indicating the type'
API, which makes it easier to set both getter and setter at once.
This was already possible before but required two calls of this
function, which wasn't intuitive:
define_accessor(name, getter, true, ...);
define_accessor(name, setter, false, ...);
Which now becomes:
define_accessor(name, getter, setter, ...);
Letting these create and return a JS::Array directly is pretty awkward
since we then need to go through the indexed properties for iteration.
Just use a MarkedValueList (i.e. Vector<Value>) for this and add a new
Array::create_from() function to turn the Vector into a returnable
Array as we did before.
This brings it a lot closer to the spec as well, which uses the
CreateArrayFromList abstract operation to do exactly this.
There's an optimization opportunity for the future here, since we know
the Vector's size we could prepare the newly created Array accordingly,
e.g. by switching to generic storage upfront if needed.
The new default return_type argument is GetOwnPropertyReturnType::All,
which returns properties with both string and symbol keys (which is also
the default for [[OwnPropertyKeys]]). This means that in some cases we
need to iterate the ordered property table twice, as we don't store
string and symbol properties separately but symbols must - there's
certainly room for (performance) improvements here. On the other hand
this makes Reflect.ownKeys() return symbol properties now :^)
Object::get_own_properties() is a bit unwieldy to use - especially as
StringOnly is about to no longer be the default value. The spec has an
abstract operation specifically for this (EnumerateObjectProperties),
so let's use that. No functionality change.
Specifically:
- Object::get_own_properties()
- Object::put_own_property()
- Object::put_own_property_by_index()
These APIs make no sense (and are inconsistent, get_own_property()
didn't have this parameter, for example) - and as expected we were
always passing in the same object we were calling the method on anyway.
