that there is one more argument (the one following the comma) and make
the candidate non-viable if the function cannot accept any argument in
that position.
llvm-svn: 82625
results for other, textual completion. For call completion, we now
produce enough information to show the function call argument that we
are currently on.
llvm-svn: 82592
members found in base classes have the same ranking as members found
in derived classes. However, we will introduce an informative note for
members found in base classes, showing (as a nested-name-specifier)
the qualification to name the base class, to make it clear which
members are from bases.
llvm-svn: 82586
-code-completion-at=filename:line:column
which performs code completion at the specified location by truncating
the file at that position and enabling code completion. This approach
makes it possible to run multiple tests from a single test file, and
gives a more natural command-line interface.
llvm-svn: 82571
notation. There is still an issue accessing field of a 'Class''s isa
in legacy code using dot field access notation (as noted in the test case)
but unrelated to this patch.
llvm-svn: 82555
opening parentheses and after each comma. We gather the set of visible
overloaded functions, perform "partial" overloading based on the set
of arguments that we have thus far, and return the still-viable
results sorted by the likelihood that they will be the best candidate.
Most of the changes in this patch are a refactoring of the overloading
routines for a function call, since we needed to separate out the
notion of building an overload set (common to code-completion and
normal semantic analysis) and then what to do with that overload
set. As part of this change, I've pushed explicit template arguments
into a few more subroutines.
There is still much more work to do in this area. Function templates
won't be handled well (unless we happen to deduce all of the template
arguments before we hit the completion point), nor will overloaded
function-call operators or calls to member functions.
llvm-svn: 82549
Several of the existing methods were identical to their respective
specializations, and so have been removed entirely. Several more 'leaf'
optimizations were introduced.
The getAsFoo() methods which imposed extra conditions, like
getAsObjCInterfacePointerType(), have been left in place.
llvm-svn: 82501
integer pointer. For now just invalidate the fields of the struct.
This addresses: <rdar://problem/7185607> [RegionStore] support invalidation of bit fields using integer assignment
llvm-svn: 82492
a nested-name-specifier that describes how to refer to that name. For
example, given:
struct Base { int member; };
struct Derived : Base { int member; };
the code-completion result for a member access into "Derived" will
provide both "member" to refer to Derived::member (no qualification needed) and
"Base::member" to refer to Base::member (qualification included).
llvm-svn: 82476
enumerators when either the user intentionally wrote a qualified name
(in which case we just use that nested-name-specifier to match
the user's code) or when this is the first "case" statement and we
need a qualified name to refer to an enumerator in a different scope.
llvm-svn: 82474
- Currently this requires us to fake an input file.
- This allows Sema to be keep all the logic for how to pull decls out of the external AST source and how to handle things like tentative definitions.
llvm-svn: 82432
"->", or "::" if we will be looking into a dependent context. It's not
wrong to use the "template" keyword, but it's to needed, either.
llvm-svn: 82307
template smarter, by taking into account which function template
parameters are deducible from the call arguments. For example,
template<typename RandomAccessIterator>
void sort(RandomAccessIterator first, RandomAccessIterator last);
will have a code-completion string like
sort({RandomAccessIterator first}, {RandomAccessIterator last})
since the template argument for its template parameter is
deducible. On the other hand,
template<class X, class Y>
X* dyn_cast(Y *Val);
will have a code-completion string like
dyn_cast<{class X}>({Y *Val})
since the template type parameter X is not deducible from the function
call.
llvm-svn: 82306
- after "using", show anything that can be a nested-name-specifier.
- after "using namespace", show any visible namespaces or namespace aliases
- after "namespace", show any namespace definitions in the current scope
- after "namespace identifier = ", show any visible namespaces or
namespace aliases
llvm-svn: 82251
look into the current scope for anything that could start a
nested-names-specifier. These results are ranked worse than any of the
results actually found in the lexical scope.
Perform a little more pruning of the result set, eliminating
constructors, __va_list_tag, and any duplication of declarations in
the result set. For the latter, implemented
NamespaceDecl::getCanonicalDecl.
llvm-svn: 82231
will provide the names of various enumerations currently
visible. Introduced filtering of code-completion results when we build
the result set, so that we can identify just the kinds of declarations
we want.
This implementation is incomplete for C++, since we don't consider
that the token after the tag keyword could start a
nested-name-specifier.
llvm-svn: 82222
when running the analyzer on real projects. We'll keep the change to
AnalysisManager.cpp in r82198 so that -fobjc-gc analyzes code
correctly in both GC and non-GC modes, although this may emit two
diagnostics for each bug in some cases (a better solution will come
later).
llvm-svn: 82201
pruning of diagnostics that may be emitted multiple times. This is
accomplished by adding FoldingSet profiling support to PathDiagnostic,
and then having BugReporter record what diagnostics have been issued.
This was motived to a serious bug introduced by moving the
'divide-by-zero' checking outside of GRExprEngine into a separate
'Checker' class. When analyzing code using the '-fobjc-gc' option, a
given function would be analyzed twice, but the second time various
"internal checks" would be disabled to avoid emitting multiple
diagnostics (e.g., "null dereference") for the same issue. The
problem is that such checks also effect path pruning and don't just
emit diagnostics. This resulted in an assertion failure involving a
real divide-by-zero in some analyzed code where we would get an
assertion failure in APInt because the 'DivZero' check was disabled
and didn't prune the logic that resulted in the divide-by-zero in the
analyzer.
The implemented solution is somewhat of a hack, and may not perform
extremely well. This will need to be cleaned up over time.
As a regression test, 'misc-ps.m' has been modified so that its tests
are run using -fobjc-gc to test this diagnostic pruning behavior.
llvm-svn: 82198
essence, code completion is triggered by a magic "code completion"
token produced by the lexer [*], which the parser recognizes at
certain points in the grammar. The parser then calls into the Action
object with the appropriate CodeCompletionXXX action.
Sema implements the CodeCompletionXXX callbacks by performing minimal
translation, then forwarding them to a CodeCompletionConsumer
subclass, which uses the results of semantic analysis to provide
code-completion results. At present, only a single, "printing" code
completion consumer is available, for regression testing and
debugging. However, the design is meant to permit other
code-completion consumers.
This initial commit contains two code-completion actions: one for
member access, e.g., "x." or "p->", and one for
nested-name-specifiers, e.g., "std::". More code-completion actions
will follow, along with improved gathering of code-completion results
for the various contexts.
[*] In the current -code-completion-dump testing/debugging mode, the
file is truncated at the completion point and EOF is translated into
"code completion".
llvm-svn: 82166
- Move CMake to using the new test runner.
- Switch Makefiles to use the lit.site.cfg.in template.
- Remove explicit --path arguments, instead this gets written into the site
configuration. This means running lit from the command line should use the
exact same configuration as is used in 'make test', assuming it can find the
site configuration file. You still need to run 'make test' (or the cmake
build target equivalent) at least once.
llvm-svn: 82160
Anders Carlsson