This simplifies several instances of metaprogramming by leveraging
[constraints and concepts from C++20][1]. This:
* gets rid of `std::enable_if` by usage of `requires`, making it more
readable and yield better compiler messages.
* uses `requires` instead of `static_assert` to enforce `TrapLabel`
typing
* simplifies all compile-time tests for validity of a given expression
* uses some standard library concepts where possible
* generalizes and simplifies `SwiftLocationExtractor`
Notice that in order to use the `std::derived_from` concept, `virtual`
inheritance had to be added to the label tags, because diamond
inheritance is a problem otherwise. That's because
`std::derived_from<T, U>` requires that `T*` be convertible to `U*`,
which is false if there are multiple non-virtual inheritance paths from
`U` to `T`. As tags never get actually instantiated, there is no runtime
performance penalty in using `virtual` inheritance.
[1]: https://en.cppreference.com/w/cpp/language/constraints
* visiting now happens in a later stage than fetching labels. While
fetching a list of entities to be visited is created, and then acted
upon in actual extraction. This partially flattens the recursive
nature of `fetchLabel` into a loop inside `SwiftVisitor::extract`.
Recursion in `fetchLabel` will only happen on labels fetched while
naming an entity (calling into `SwiftMangler`).
* The choice whether to name a declaration or type has been moved from
the translators to `SwiftMangler`. Acting on this choice is contained
in `SwiftDispatcher::createLabel`.
* The choice whether to emit a body of a declaration has been moved from
`DeclTranslator` to the dispatcher. This choice is also contained in
`SwiftDispatcher::createLabel`.
* The simple functionality of the `LabelStore` has been moved to the
`SwiftDispatcher` as well.
It's not much cleaner due to arithmetic to convert truncating division to a ceiling, but has two advantages:
1. It doesn't suffer from rounding issues with large TRAP labels. This is largely theoretical, but does let us handle `undefined` uniformly.
2. It should be much faster (using LZCNT/BSR instead of floating point arithmetic). This is probably not a performance bottleneck, so *shrug*.
That class was meant to allow aggregate initialization of generated
C++ entries having the label `id` as first argument.
As aggregate initialization turned out to be undesirable (names of
fields are not explicit, and `{}` must be inserted for empty
superclasses), this commit removes it and disallows aggregate
initialization altogether by defining empty constructors for generated
classes.
This allows to avoid bypassing label type correcness in the extractor,
and allows to independently resolve TBD extractions, as with this
approach TBD nodes do have the correctly typed trap label. The TBD
status is now a predicate on the QL side.
This requires:
* a default visit using the correct type, which is achieved via macro
metaprogramming in `VisitorBase.h`, following the way
`swift::ASTVisitor` is programmed
* a mapping from labels to corresponding binding trap entries. The
functor is defined in `TrapTagTraits.h` and instantiated in generated
`TrapEntries.h`
* Binding trap entries for TBD unknown entities must not have any other
field than the `id` (after all, we are supposed to not extract them
yet). This is why all unextracted fields in `schema.yml` have been
commented out, and will be uncommentend when visitors are added
This turned out easier than expected previously. `llvm::PointerUnion`
was also considered, which would have less memory footprint, but it
would require more effort as it is lacking the same implicit conversions
and operators that `std::variant` provides.
Also renamed `ToTag<E>` to `TrapTagOf<E>` and introduced a derived
convenience functor `TrapLabelOf<E>`.
These changes are required to allow a new type-safe approach to TBD
nodes, that will come in a separate commit.
This introduces:
* the possibility to add properties to the root `Element`
* a functor taking tags to the corresponding binding trap entry
* `hasProp()` methods for optional properties in QL
* `getPrimaryQlClass()` method
This checks in the trapgen script generating trap entries in C++.
The codegen suite has been slightly reorganized, moving the templates
directory up one level and chopping everything into smaller bazel
packages. Running tests is now done via
```
bazel run //swift/codegen/test
```
With respect to the PoC, the nested `codeql::trap` namespace has been
dropped in favour of a `Trap` prefix (or suffix in case of entries)
within the `codeql` namespace. Also, generated C++ code is not checked
in in git any more, and generated during build. Finally, labels get
printed in hex in the trap file.
`TrapLabel` is for the moment only default-constructible, so only one
single label is possible. `TrapArena`, that is responsible for creating
disjoint labels will come in a later commit.
