Commit Graph

710 Commits

Author SHA1 Message Date
Rintaro Ishizaki
3b42894f13 [Parse] Fix parsing three-version-components in #if
* Narrow allowance of 3+ components numeric literal to condition part of the
  directive.
* Allow 3+ components in '#if' directive in decl list position as well.
2017-02-16 11:35:51 +09:00
Rintaro Ishizaki
cfe742d1eb [Parse] Minor improvements in conditional compilation block parsing
* Don't emit duplicated 'expected #else or #endif at end of conditional
  compilation block' error.

    class Foo {
      #if true
        func foo() {}
    [EOF]

* Improve error message when seeing '}' in config block.

    class Foo {
    #if true
        func foo();
    } // error: unexpected '}' in conditional compilation block
    #else
    #endif
2017-02-16 11:35:51 +09:00
Rintaro Ishizaki
49d134f1a9 [Parse] Stop lying about QuestionLoc of implicit OptionalTypeRepr (#7384)
In conditional statement let/var patterns.
Set invalid location instead.
2017-02-11 13:39:51 +09:00
Jacob Bandes-Storch
e1fd8aa6c7 [Parse] Fix crash in conditional compilation parsing (#7331) 2017-02-08 23:10:26 -08:00
Rintaro Ishizaki
ff826e3491 [Parse] Limit max nesting level of StructureMarkerRAII (#7330)
The threshold is 256, for now.
2017-02-09 15:04:09 +09:00
Jacob Bandes-Storch
c98e515734 [QoI] Improvements to function call & closure diagnostics (#7224) 2017-02-07 17:36:11 -08:00
Doug Gregor
81daa1f53b [Type checker] Be more tolerant of broken initializer declarations. 2017-02-07 15:22:50 -08:00
Slava Pestov
4f4e6f6ebf Sema: Remove another workaround and fuzz harder
We used to drop the entire generic parameter list if one of the
entries failed to parse. This caused a problem where the generic
parameters were still available for name lookup, so they had
to be special-cased since there's no generic environment set up
in this case.

Now, keep the parts of the generic parameter list around that
parsed successfully.

When I first made the change, almost a hundred crashers regressed;
now all the underlying issues have been fixed.

The result is that in addition to removing a crappy hack we get
some more mileage out of the compiler_crashers, because stuff like
this now builds a generic environment:

class S<T{...}
2017-02-03 17:02:04 -08:00
Jacob Bandes-Storch
3e7e923e6d [Parse] Reject trailing closures on literals (#7202)
`1 { }` was parsed as a call expression with a trailing closure. This made the diagnostics for `var x = 1 { get { ... } }` extremely bad. Resolves SR-3671.
2017-02-02 10:32:47 -08:00
Slava Pestov
d880053dd5 Merge pull request #7050 from rintaro/parse-diag-consecutiveids
[Parse] Improve diagnostics for consecutive identifiers
2017-01-31 21:19:19 -08:00
Slava Pestov
71cf245701 Merge pull request #7023 from KingOfBrian/bugfix/SR-2115
Generate unused variable warnings in top level statements
2017-01-29 20:25:46 -08:00
Brian King
0c57aebfea Fix unit tests 2017-01-26 10:04:41 -05:00
Rintaro Ishizaki
9b4f549943 [Parse] Improve diagnostics for consecutive identifiers 2017-01-26 15:37:19 +09:00
Brian King
edc193efa0 Fix warnings in top level statements 2017-01-24 22:53:21 -05:00
Slava Pestov
a67f4bdff4 Merge pull request #6948 from matthewcarroll/SR-3599-Better-Recovery-For-Decl-Names
[DiagnosticsQoI] SR-3599: Better recovery for decls with consecutive identifiers
2017-01-23 17:37:07 -08:00
Jordan Rose
dec7f9fe7e Merge pull request #6863 from matthewcarroll/SR-3600-Better-recovery-for-naming-an-initializer-deinitializer-or-subscript
[DiagnosticsQoI] SR-3600: Better recovery for trying to name an initializer, deinitializer, or subscript
2017-01-23 16:30:50 -08:00
Matthew Carroll
f8f147ecf9 [DiagnosticsQoI] SR-3600: Better recovery for trying to name an initializer, deinitializer, or subscript
- Restrict this diagnostic to identifiers that are followed by a left paren.
2017-01-21 11:32:09 -05:00
Slava Pestov
e8bd7020fa Merge pull request #6855 from rintaro/parse-ifconfig-invalid-binary
[Parse] Disallow unsupported condition expression in #if directive
2017-01-20 20:27:45 -08:00
Matthew Carroll
93a7a9f80d [DiagnosticsQoI] SR-3599: Better recovery for decls with consecutive identifiers
Add diagnostics to fix decls with consecutive identifiers. This applies to
types, properties, variables, and enum cases. The diagnostic adds a camel-cased option if it is different than the first option.

https://bugs.swift.org/browse/SR-3599
2017-01-19 23:44:33 -05:00
Matthew Carroll
7882ddcf40 [DiagnosticsQoI] SR-3600: Better recovery for trying to name an initializer, deinitializer, or subscript
Add a diagnostic to remove the name of an initializer, deinitializer, or subscript. The identifier token is consumed and skipped to prevent the parser from emitting additional error messages.

Add tests to verify that the name is removed from initializers, deinitializers, and subscripts declared with a name.
2017-01-17 17:21:56 -05:00
Rintaro Ishizaki
b950f06f3e [Parse] Disallow unsupported condition expression in #if directive
In non-Swift3 mode.

Previously:

  #if FOO = false
  #elseif FOO ? false : true
  #endif

were silently accepted.
i.e. '= false' and '? false : true' were silently ignored.
2017-01-17 22:11:11 +09:00
Rintaro Ishizaki
055ee48908 [Parse] Split Decl list parsing from parseList()
Parsing declaration list (e.g. member list of nominal decl) is very
different from comma separated list, because it's elements are separated with
new-line or semi-colon. There's no good reason to consolidate them.

Also, declaration list in 'extension' or inside of decl '#if' didn't
emit diagnostics for consecutive declarations on a line.

    class C {
    #if true
      var value: Int = 42 func foo() {}
    #endif
    }
    extension C {
      func bar() {} subscript(i: Int) -> Int {
        return 24
      }
    }

This change consolidates declaration list parsing for
members of nominal decls, extensions, and inside of '#if'.

In addition, removed unnecessary property 'TrailingSemiLoc' from decl.
2017-01-17 17:19:36 +09:00
Rintaro Ishizaki
70d53525e7 [Parse] Add testcases for tokenizing SIL keywords as identifiers in non-SIL mode 2017-01-17 03:10:50 +09:00
Slava Pestov
48ef1ad00d Merge pull request #5721 from rintaro/parse-functy
[Parse] Minor code improvement in function-type parsing
2017-01-12 19:22:34 -08:00
Jacob Bandes-Storch
49f5251c73 [QoI] offer typo correction for platform conditionals 2017-01-12 09:07:53 -08:00
Jacob Bandes-Storch
4a16880103 [QoI] Add diagnostic for unused literal values
Fixes SR-3522.
2017-01-10 20:42:36 -08:00
Robert Widmann
d2570cb964 Restore local type declarations list 2017-01-06 16:16:39 -07:00
practicalswift
30a88d38e6 [gardening] Fix recently introduced typos 2017-01-06 21:16:02 +01:00
Robert Widmann
dfa41d625b Optimize Condition Resolution (#6279)
* Pack the bits for IfConfigDecls into Decl

* Don't open symbols into a module when evaluating canImport statements

The module loaders now have API to check whether a given module can be
imported without importing the referenced module.  This provides a
significant speed boost to condition resolution and no longer
introduces symbols from the referenced module into the current context
without the user explicitly requesting it.

The definition of ‘canImport’ does not necessarily mean that a full
import without error is possible, merely that the path to the import is
visible to the compiler and the module is loadable in some form or
another.

Note that this means this check is insufficient to guarantee that you
are on one platform or another.  For those kinds of checks, use
‘os(OSNAME)’.
2017-01-05 12:08:54 -07:00
Slava Pestov
8f96606cb2 Sema: Remove TupleToScalar conversion, which did nothing useful except crash
As far as I can tell there's no way to trigger this from valid code.
2017-01-03 23:00:07 -08:00
Joe Groff
1889fde228 Resolve type(of:) by overload resolution rather than parse hackery.
`type(of:)` has behavior whose type isn't directly representable in Swift's type system, since it produces both concrete and existential metatypes. In Swift 3 we put in a parser hack to turn `type(of: <expr>)` into a DynamicTypeExpr, but this effectively made `type(of:)` a reserved name. It's a bit more principled to put `Swift.type(of:)` on the same level as other declarations, even with its special-case type system behavior, and we can do this by special-casing the type system we produce during overload resolution if `Swift.type(of:)` shows up in an overload set. This also lays groundwork for handling other declarations we want to ostensibly behave like normal declarations but with otherwise inexpressible types, viz. `withoutActuallyEscaping` from SE-0110.
2016-12-22 16:28:31 -08:00
Rintaro Ishizaki
808c05c6b1 Merge pull request #6381 from rintaro/sema-pattern-specialized-call
[TypeChecker] Don't convert specialzed call into EnumElement pattern
2016-12-22 11:02:48 +09:00
practicalswift
9e60bdbffa [gardening] Fix typos. 2016-12-21 21:33:30 +01:00
Slava Pestov
3f7a72eb84 Sema: Clean up diagnostic logic when deriving RawRepresentable conformance on enums 2016-12-21 14:20:26 -05:00
Rintaro Ishizaki
509db744f1 [Lexer] Disallow '$' as a start of identifier, special handle '$' (#6004)
Allowing 'let `$0` = 1', introduced in 6accc598 was not intentional.
2016-12-21 11:31:58 +09:00
Rintaro Ishizaki
6d477a557f [Tests] Add FIXME test cases for enum-element-pattern with sub pattern 2016-12-21 02:22:48 +09:00
Rintaro Ishizaki
9f43e456f4 [TypeChecker] Add diagnostic with fix-it for specialized enum case pattern in Swift3 mode. 2016-12-21 02:22:20 +09:00
Rintaro Ishizaki
80721e9c30 [TypeChecker] Don't convert specialzed call into EnumElement pattern 2016-12-20 20:21:48 +09:00
Robert Widmann
9fd7b5a29c Validate the entire #if configuration
Being lazy here just means we don’t validate the entire if
configuration.  We need to be able to emit diagnostics even in
condition clauses with indeterminate expressions.
2016-12-19 02:32:47 -05:00
Robert Widmann
a7957a199b Walk into top level pattern binding decls to resolve conditions 2016-12-15 04:50:02 -05:00
Robert Widmann
cededef0d6 Add condition resolution as a new phrase post-parse
An unfortunately necessary thing to delay defrosting function bodies as
long as we can.
2016-12-14 15:39:19 -05:00
Robert Widmann
72beb9d80d Extract common code into StmtTransformer 2016-12-14 14:59:47 -05:00
Robert Widmann
a060eb5aca [SE-0075] Transfer the power of config resolution to Namebinding
This completely removes Parse’s ability to make any judgement calls
about compilation conditions, instead the parser-relevant parts of
‘evaluateConditionalCompilationExpr’ have been moved into
‘classifyConditionalCompilationExpr’ where they exist to make sure only
decls that we want to parse actually parse later.

The condition-evaluation parts have been moved into NameBinding in the
form of a Walker that evaluates and collapses IfConfigs.  This walker
is meant as an homage to PlaygroundLogger.  It should probably be
factored out into a common walker at some point in the future.
2016-12-14 14:59:47 -05:00
Robin Kunde
ab415e6b3e [Parse] Improve parser diagnostics for keyword-as-identifer errors (#6045)
Instead of the simple "expected identifier in declaration", the error will now read "keyword '%' cannot be used as an identifier here", and will be accompanied by a note suggesting escaping the keyword with backticks, as well as a fixit.

https://bugs.swift.org/browse/SR-3167
2016-12-13 16:12:53 -08:00
Rintaro Ishizaki
42ebd38c29 [Parse] Skip non-case-label statements in switch (#6215) 2016-12-13 03:37:09 +09:00
Rintaro Ishizaki
9eee4adc91 [Parse] Parameter list parsing cleanup
* Removed `parseConstructorArguments()`, unified with
  `parseSingleParameterClause()`.
* Use `parseSingleParameterClause()` from `parseFunctionSignature()`, so
  that we can share the recovery code.
* Removed `isFirstParameterClause` parameter from `mapParsedParameters`,
  because it's predictable from `paramContext`.
2016-12-07 23:20:30 +09:00
Rintaro Ishizaki
ec53b6885f [Parse] Let #if diretives be parsed in parseBraceItems()
parseBraceItems() has specific logic for pasing conditional compilation blocks.
Withoutout that, decralations in the block cannot be propagated to outside.
For instance:

  FOO: #if true
  func foo() {}
  #endif
  foo() // error: use of unresolved identifier 'foo'
2016-12-06 22:28:53 +09:00
Rintaro Ishizaki
54efbbbf87 [Parse] Let #sourceLocation be parsed in parseBraceItems()
Returning 'makeParserSuccess()' as a 'ParserResult<Stmt>' causes assertion error.

  Assertion `Status.isError()' failed
2016-12-06 22:28:53 +09:00
David Farler
330c2d96e6 Make the lexer UTF-8 RFC 3629 correct re: prefix octets
RFC 2279 states that, in UTF-8:
"The octet values FE and FF never appear."

RFC 3629 states that, in UTF-8:
"The octet values C0, C1, F5 to FF never appear."

Generalize the check to advance past invalid starting bytes for
a UTF-8 sequence to fix a crash in the lexer.
2016-12-05 17:21:17 -08:00
Rintaro Ishizaki
2a2da1f7e7 [Parse] Add FIXME for bad recovery for incomplete function type 2016-11-30 23:09:38 +09:00