Rework the lazy function body parsing mechanism to use the
request-evaluator, so that asking for the body of a function will
initiate parsing. Clean up a number of callers to
AbstractFunctionDecl::getBody() that don't actually need the body, so
we don't perform unnecessary parsing.
This change does not delay parsing of function bodies in the general
case; rather, it sets up the infrastructure to always delay parsing of
function bodies.
Rework the lazy function body parsing mechanism to use the
request-evaluator, so that asking for the body of a function will
initiate parsing. Clean up a number of callers to
AbstractFunctionDecl::getBody() that don't actually need the body, so
we don't perform unnecessary parsing.
This change does not delay parsing of function bodies in the general
case; rather, it sets up the infrastructure to always delay parsing of
function bodies.
Lazy parsing for the members of nominal types and extensions depends
only on information already present in
`IterableDeclContext`. Eliminate the use of PersistentParserState as
an intermediary and have the member-parsing request construct a new
`Parser` instance itself to handle parsing. Make this possible even
for ill-formed nominal types/extensions to simplify the code path.
Eliminate `LazyMemberParser` and all of its uses, because it was only
present for lazy member parsing, which no longer needs it.
Ensure that lazy parsing of the members of nominal type definitions
and extensions is handled through a request. Most of the effort here
is in establishing a new request zone for parser requests.
To properly delay parsing type and extension bodies we need to know
which ones might contain nested operator and class definitions, since
they must be known upfront when building the global operator lookup
and AnyObject dispatch lookup tables, respectively.
To guess if the type contains operator definitions, we look for the
'func' keyword followed by an operator token.
To guess if the type contains class definitions, we look for the
'class' keyword.
For now, this information is recorded but not used. Subsequent commits
will make use of this information to delay parsing in more cases.
File: [swift/lib/Parse/PersistentParserState.cpp]
Observing this code snipped closely in function, [void PersistentParserState::parseAllDelayedDeclLists()]:
```
for (auto &P: DelayedDeclListStates) {
AllDelayed.push_back(P.first);
}
```
We observe that 'AllDelayed' gets maximum of DelayedDeclListStates.size() elements pushed into it.
So it is better to reserve 'vector AllDelayed' to that max size, to save reallocation cost.
I believe we can use size() on [DelayedDeclListStates], which is of type: ```llvm::DenseMap<IterableDeclContext *, std::unique_ptr<DelayedDeclListState>>``` because of the ```DenseMapIterator``` implementation of size().
[http://llvm.org/doxygen/DenseMap_8h_source.html#l00126]
Now that, parsing accessor body is done by 'parseAbstractFunctionBody()'
which automatically handles delayed parsing by `delayFunctionBodyParsing()`.
So we no longer use delayed parsing facilities specific for accessor body.
completion inside computed properties.
Adding tests for willSet/didSet uncovered some crashes while doing code
completion (see FIXMEs), and I will investigate these next.
Partially addresses rdar://15849262
Swift SVN r14338
This allows us to show generic parameters in:
struct S<T> {
func f(a: #^A^#
}
And show the type Z in:
struct S {
func f(a: #^A^#
typealias Z = Int
}
Swift SVN r7216
-Introduce PersistentParserState to represent state persistent among multiple parsing passes.
The advantage is that PersistentParserState is independent of a particular Parser or Lexer object.
-Use PersistentParserState to keep information about delayed function body parsing and eliminate parser-specific
state from the AST (ParserTokenRange).
-Introduce DelayedParsingCallbacks to abstract out of the parser the logic about which functions should be delayed
or skipped.
Many thanks to Dmitri for his valuable feedback!
Swift SVN r6580
