While adding spec comments to PerformEval, I noticed we were missing
multiple steps.
Namely, these were:
- Checking if the host will allow us to compile the string
(allowing LibWeb to perform CSP for eval)
- The parser's initial state depending on the environment around us
on direct eval:
- Allowing new.target via eval in functions
- Allowing super calls and super properties via eval in classes
- Disallowing the use of the arguments object in class field
initializers at eval's parse time
- Setting ScriptOrModule of eval's execution context
The spec allows us to apply the additional parsing steps in any order.
The method I have gone with is passing in a struct to the parser's
constructor, which overrides the parser's initial state to (dis)allow
the things stated above from the get-go.
The same expression is not allowed to contain both the
logical && and || operators, and the coalescing ?? operator.
This patch changes how "forbidden" tokens are handled, using a
finite set instead of an Vector. This supports much more efficient
merging of the forbidden tokens when propagating forward, and
allowing the return of forbidden tokens to parent contexts.
This patch makes check_identifier_name_for_assignment_validity()
take a FlyString instead of a StringView. We then exploit this by
passing FlyString in more places via flystring_value().
This gives a ~1% speedup when parsing the largest Discord JS file.
When parsing identifiers, we ultimately want to sink the token string
into a FlyString anyway, and since Token may have a FlyString already
inside it, this allows us to bypass the costly FlyString(StringView).
This gives a ~3% speedup when parsing the largest Discord JS file.
Everyone who calls this already has a FlyString, so we were doing *way*
more work by pessimizing it to a StringView.
This gives a ~2% speedup when parsing the largest Discord JS file.
This removes a number of vm.exception() checks which are now caught
directly by TRY. Make use of these checks in
{Global, Eval}DeclarationInstantiation and while we're here add spec
comments.
Because we can have arbitrary in- and export names with strings we can
have '*' and '' which means using '*' as an indicating namespace imports
failed / behaved incorrectly for string imports '*'.
We now use more specific types to indicate these special states instead
of these 'magic' string values.
Do note that 'default' is not actually a magic string value but one
specified by the spec. And you can in fact export the default value by
doing: `export { 1 as default }`.
The big changes are:
- Allow strings as Module{Export, Import}Name
- Properly track declarations in default export statements
However, the spec is a little strange in that it allows function and
class declarations without a name in default export statements.
This is quite hard to fully implement without rewriting more of the
parser so for now this behavior is emulated by faking things with
function and class expressions. See the comments in
parse_export_statement for details on the hacks and where it goes wrong.
The three major changes are:
- Parsing parameters, the function body, and then the full assembled
function source all separately. This is required by the spec, as
function parameters and body must be valid each on their own, which
cannot be guaranteed if we only ever parse the full function.
- Returning an ECMAScriptFunctionObject instead of a FunctionExpression
that needs to be evaluated separately. This vastly simplifies the
{Async,AsyncGenerator,Generator,}Function constructor implementations.
Drop '_node' from the function name accordingly.
- The prototype is now determined via GetPrototypeFromConstructor and
passed to OrdinaryFunctionCreate.
The parentheses are dealt with outside this function, so we shouldn't
use the (non)existence of one as the condition for consuming another
comma and then parameter. Just check for a comma instead. This becomes
relevant when parsing standalone function parameters as per the spec in
the CreateDynamicFunction AO.
The curly braces are not part of the FunctionBody production. This
becomes relevant when parsing a standalone function body as per the spec
in the CreateDynamicFunction AO.
This includes:
- Parsing proper LabelledStatements with try_parse_labelled_statement()
- Removing LabelableStatement
- Implementing the LoopEvaluation semantics via loop_evaluation() in
each IterationStatement subclass; and IterationStatement evaluation
via {For,ForIn,ForOf,ForAwaitOf,While,DoWhile}Statement::execute()
- Updating ReturnStatement, BreakStatement and ContinueStatement to
return the appropriate completion types
- Basically reimplementing TryStatement and SwitchStatement according to
the spec, using completions
- Honoring result completion types in AsyncBlockStart and
OrdinaryCallEvaluateBody
- Removing any uses of the VM unwind mechanism - most importantly,
VM::throw_exception() now exclusively sets an exception and no longer
triggers any unwinding mechanism.
However, we already did a good job updating all of LibWeb and userland
applications to not use it, and the few remaining uses elsewhere don't
rely on unwinding AFAICT.
Although those are the only valid options parse_primary_expression is
sometimes called when only an expression is valid which means it did not
check match_expression and might fail the now removed VERIFY.
This allows us to only perform checks like export bindings existing only
for modules. Also this makes it easier to set strict and other state
variables with TemporaryChanges.
In static init blocks 'await' cannot be used. Note that this does not
cover all the cases since the parser currently cannot distinguish
between expressions within parenthesis and direct expressions.
This commit adds support for the most bare bones version of async
functions, support for async generator functions, async arrow functions
and await expressions are TODO.
