Merge pull request #22952 from nkcsgexi/cherry-pick-ide

[5.1] cherry pick recent sourcekitd/IDE fixes

Author: nkcsgexi

include/swift/AST/DiagnosticsCommon.def

@@ -50,6 +50,12 @@ ERROR(not_implemented,none,
 ERROR(error_opening_output,none,
       "error opening '%0' for output: %1", (StringRef, StringRef))
 
+ERROR(cannot_find_group_info_file,none,
+      "cannot find group info file at path: '%0'", (StringRef))
+
+ERROR(cannot_parse_group_info_file,none,
+      "cannot parse group info file at path: '%0'", (StringRef))
+
 ERROR(error_no_group_info,none,
       "no group info found for file: '%0'", (StringRef))
 

include/swift/Sema/IDETypeChecking.h

@@ -161,6 +161,28 @@ namespace swift {
     bool shouldPrintRequirement(ExtensionDecl *ED, StringRef Req);
     bool hasMergeGroup(MergeGroupKind Kind);
   };
+
+  /// Reported type for an expression. This expression is represented by offset
+  /// length in the source buffer;
+  struct ExpressionTypeInfo {
+
+    /// The start of the expression;
+    uint32_t offset;
+
+    /// The length of the expression;
+    uint32_t length;
+
+    /// The start of the printed type in a separately given string buffer.
+    uint32_t typeOffset;
+
+    /// The length of the printed type
+    uint32_t typeLength;
+  };
+
+  /// Collect type information for every expression in \c SF; all types will
+  /// be printed to \c OS.
+  ArrayRef<ExpressionTypeInfo> collectExpressionType(SourceFile &SF,
+    std::vector<ExpressionTypeInfo> &scratch, llvm::raw_ostream &OS);
 }
 
 #endif

lib/IDE/IDETypeChecking.cpp

@@ -23,6 +23,8 @@
 #include "swift/AST/NameLookup.h"
 #include "swift/AST/ProtocolConformance.h"
 #include "swift/Sema/IDETypeChecking.h"
+#include "swift/IDE/SourceEntityWalker.h"
+#include "swift/Parse/Lexer.h"
 
 using namespace swift;
 
@@ -577,3 +579,84 @@ collectDefaultImplementationForProtocolMembers(ProtocolDecl *PD,
   for (auto *IP : PD->getInheritedProtocols())
     HandleMembers(IP->getMembers());
 }
+
+/// This walker will traverse the AST and report types for every expression.
+class ExpressionTypeCollector: public SourceEntityWalker {
+  SourceManager &SM;
+  unsigned int BufferId;
+  std::vector<ExpressionTypeInfo> &Results;
+
+  // This is to where we print all types.
+  llvm::raw_ostream &OS;
+
+  // Map from a printed type to the offset in OS where the type starts.
+  llvm::StringMap<uint32_t> TypeOffsets;
+
+  // This keeps track of whether we have a type reported for a given
+  // [offset, length].
+  llvm::DenseMap<unsigned, llvm::DenseSet<unsigned>> AllPrintedTypes;
+
+  bool shouldReport(unsigned Offset, unsigned Length, Expr *E) {
+    // We shouldn't report null types.
+    if (E->getType().isNull())
+      return false;
+
+    // If we have already reported types for this source range, we shouldn't
+    // report again. This makes sure we always report the outtermost type of
+    // several overlapping expressions.
+    auto &Bucket = AllPrintedTypes[Offset];
+    return Bucket.find(Length) == Bucket.end();
+  }
+
+  // Find an existing offset in the type buffer otherwise print the type to
+  // the buffer.
+  uint32_t getTypeOffsets(StringRef PrintedType) {
+    auto It = TypeOffsets.find(PrintedType);
+    if (It == TypeOffsets.end()) {
+      TypeOffsets[PrintedType] = OS.tell();
+      OS << PrintedType;
+    }
+    return TypeOffsets[PrintedType];
+  }
+
+public:
+  ExpressionTypeCollector(SourceFile &SF, std::vector<ExpressionTypeInfo> &Results,
+    llvm::raw_ostream &OS): SM(SF.getASTContext().SourceMgr),
+                            BufferId(*SF.getBufferID()),
+                            Results(Results), OS(OS) {}
+  bool walkToExprPre(Expr *E) override {
+    if (E->getSourceRange().isInvalid())
+      return true;
+    CharSourceRange Range =
+      Lexer::getCharSourceRangeFromSourceRange(SM, E->getSourceRange());
+    unsigned Offset = SM.getLocOffsetInBuffer(Range.getStart(), BufferId);
+    unsigned Length = Range.getByteLength();
+    if (!shouldReport(Offset, Length, E))
+      return true;
+    // Print the type to a temporary buffer.
+    SmallString<64> Buffer;
+    {
+      llvm::raw_svector_ostream OS(Buffer);
+      E->getType()->getRValueType()->reconstituteSugar(true)->print(OS);
+      // Ensure the end user can directly use the char*
+      OS << '\0';
+    }
+
+    // Add the type information to the result list.
+    Results.push_back({Offset, Length, getTypeOffsets(Buffer.str()),
+      static_cast<uint32_t>(Buffer.size()) - 1});
+
+    // Keep track of that we have a type reported for this range.
+    AllPrintedTypes[Offset].insert(Length);
+    return true;
+  }
+};
+
+ArrayRef<ExpressionTypeInfo>
+swift::collectExpressionType(SourceFile &SF,
+                             std::vector<ExpressionTypeInfo> &Scratch,
+                             llvm::raw_ostream &OS) {
+  ExpressionTypeCollector Walker(SF, Scratch, OS);
+  Walker.walk(SF);
+  return Scratch;
+}

lib/IDE/Refactoring.cpp

@@ -1909,8 +1909,7 @@ class ExpandableAssignTernaryExprInfo: public ExpandableTernaryExprInfo {
   IfExpr *getIf() {
     if (!Assign)
       return nullptr;
-
-    return dyn_cast<IfExpr>(Assign->getSrc());
+    return dyn_cast_or_null<IfExpr>(Assign->getSrc());
   }
 
   SourceRange getNameRange() {

lib/Serialization/SerializeDoc.cpp

@@ -38,6 +38,7 @@ using pFileNameToGroupNameMap = std::unique_ptr<FileNameToGroupNameMap>;
 
 namespace {
 class YamlGroupInputParser {
+  ASTContext &Ctx;
   StringRef RecordPath;
   static constexpr const char * const Separator = "/";
 
@@ -86,12 +87,20 @@ class YamlGroupInputParser {
   }
 
 public:
-  YamlGroupInputParser(StringRef RecordPath): RecordPath(RecordPath) {}
+  YamlGroupInputParser(ASTContext &Ctx, StringRef RecordPath):
+    Ctx(Ctx), RecordPath(RecordPath) {}
 
   FileNameToGroupNameMap* getParsedMap() {
     return AllMaps[RecordPath].get();
   }
 
+  bool diagnoseGroupInfoFile(bool FileMissing = false) {
+    Ctx.Diags.diagnose(SourceLoc(),
+      FileMissing ? diag::cannot_find_group_info_file:
+      diag::cannot_parse_group_info_file, RecordPath);
+    return true;
+  }
+
   // Parse the Yaml file that contains the group information.
   // True on failure; false on success.
   bool parse() {
@@ -102,32 +111,31 @@ class YamlGroupInputParser {
 
     auto Buffer = llvm::MemoryBuffer::getFile(RecordPath);
     if (!Buffer) {
-      // The group info file does not exist.
-      return true;
+      return diagnoseGroupInfoFile(/*Missing File*/true);
     }
     llvm::SourceMgr SM;
     llvm::yaml::Stream YAMLStream(Buffer.get()->getMemBufferRef(), SM);
     llvm::yaml::document_iterator I = YAMLStream.begin();
     if (I == YAMLStream.end()) {
       // Cannot parse correctly.
-      return true;
+      return diagnoseGroupInfoFile();
     }
     llvm::yaml::Node *Root = I->getRoot();
     if (!Root) {
       // Cannot parse correctly.
-      return true;
+      return diagnoseGroupInfoFile();
     }
 
     // The format is a map of ("group0" : ["file1", "file2"]), meaning all
     // symbols from file1 and file2 belong to "group0".
     auto *Map = dyn_cast<llvm::yaml::MappingNode>(Root);
     if (!Map) {
-      return true;
+      return diagnoseGroupInfoFile();
     }
     pFileNameToGroupNameMap pMap(new FileNameToGroupNameMap());
     std::string Empty;
     if (parseRoot(*pMap, Root, Empty))
-      return true;
+      return diagnoseGroupInfoFile();
 
     // Save the parsed map to the owner.
     AllMaps[RecordPath] = std::move(pMap);
@@ -168,7 +176,7 @@ class DeclGroupNameContext {
       StringRef FullPath =
           Ctx.SourceMgr.getIdentifierForBuffer(PathOp.getValue());
       if (!pMap) {
-        YamlGroupInputParser Parser(RecordPath);
+        YamlGroupInputParser Parser(Ctx, RecordPath);
         if (!Parser.parse()) {
 
           // Get the file-name to group map if parsing correctly.

test/IDE/Inputs/ExprType.swift

@@ -0,0 +1,22 @@
+
+func foo() -> Int  { return 1 }
+
+func bar(f: Float) -> Float { return bar(f: 1) }
+
+protocol P {}
+
+func fooP(_ p: P) { fooP(p) }
+
+class C {}
+
+func ArrayC(_ a: [C]) {
+	_ = a.count
+	_ = a.description.count.advanced(by: 1).description
+}
+
+struct S {
+  let val = 4
+}
+func DictS(_ a: [Int: S]) {
+  _ = a[2]?.val.advanced(by: 1).byteSwapped
+}

test/IDE/expr_type.swift

@@ -0,0 +1,8 @@
+// RUN: %target-swift-ide-test -print-expr-type -source-filename %S/Inputs/ExprType.swift -swift-version 5 | %FileCheck %s
+
+// CHECK: func foo() -> Int { return <expr type:"Int">1</expr> }
+// CHECK: func bar(f: Float) -> Float { return <expr type:"Float"><expr type:"(Float) -> Float">bar</expr><expr type:"(f: Float)">(f: <expr type:"Float">1</expr>)</expr></expr> }
+// CHECK: func fooP(_ p: P) { <expr type:"()"><expr type:"(P) -> ()">fooP</expr><expr type:"(P)">(<expr type:"P">p</expr>)</expr></expr> }
+// CHECK: <expr type:"()"><expr type:"Int">_</expr> = <expr type:"Int"><expr type:"[C]">a</expr>.count</expr></expr>
+// CHECK: <expr type:"()"><expr type:"String">_</expr> = <expr type:"String"><expr type:"Int"><expr type:"(Int) -> Int"><expr type:"Int"><expr type:"String"><expr type:"[C]">a</expr>.description</expr>.count</expr>.<expr type:"(Int) -> (Int) -> Int">advanced</expr></expr><expr type:"(by: Int)">(by: <expr type:"Int">1</expr>)</expr></expr>.description</expr></expr>
+// CHECK: <expr type:"()"><expr type:"Int?">_</expr> = <expr type:"Int?"><expr type:"Int"><expr type:"(Int) -> Int"><expr type:"Int"><expr type:"S"><expr type:"S?"><expr type:"[Int : S]">a</expr><expr type:"(Int)">[<expr type:"Int">2</expr>]</expr></expr>?</expr>.val</expr>.<expr type:"(Int) -> (Int) -> Int">advanced</expr></expr><expr type:"(by: Int)">(by: <expr type:"Int">1</expr>)</expr></expr>.byteSwapped</expr></expr>

test/Serialization/Inputs/corrupted_group_info.json

@@ -0,0 +1,3 @@
+{
+	corrupteddata
+}

test/Serialization/group_info_diags.swift

@@ -0,0 +1,10 @@
+// RUN: %empty-directory(%t)
+// RUN: not %target-swift-frontend -emit-module %s -module-name HasArray -o %t -module-cache-path %t/mcp -group-info-path %t/nonexist.json -emit-module-doc &> %t/result.txt
+// RUN: %FileCheck %s -check-prefix=MISSING < %t/result.txt
+// RUN: not %target-swift-frontend -emit-module %s -module-name HasArray -o %t -module-cache-path %t/mcp -group-info-path %S/Inputs/corrupted_group_info.json -emit-module-doc &> %t/result.txt
+// RUN: %FileCheck %s -check-prefix=CORRUPTED < %t/result.txt
+
+public protocol P {}
+
+// MISSING: error: cannot find group info file at path
+// CORRUPTED: error: cannot parse group info file at path

test/SourceKit/ExpressionType/basic.swift

@@ -0,0 +1,34 @@
+func foo() -> Int  { return 1 }
+
+func bar(f: Float) -> Float { return bar(f: 1) }
+
+protocol P {}
+
+func fooP(_ p: P) { fooP(p) }
+
+class C {}
+
+func ArrayC(_ a: [C]) {
+	_ = a.count
+	_ = a.description.count.advanced(by: 1).description
+}
+
+struct S {
+  let val = 4
+}
+func DictS(_ a: [Int: S]) {
+  _ = a[2]?.val.advanced(by: 1).byteSwapped
+}
+
+// RUN: %sourcekitd-test -req=collect-type %s -- %s | %FileCheck %s
+// CHECK: (183, 202): Int
+// CHECK: (183, 196): String
+// CHECK: (183, 184): [C]
+// CHECK: (203, 211): (Int) -> (Int) -> Int
+// CHECK: (211, 218): (by: Int)
+// CHECK: (216, 217): Int
+// CHECK: (257, 258): Int
+// CHECK: (291, 332): ()
+// CHECK: (291, 292): Int?
+// CHECK: (295, 332): Int?
+// CHECK: (295, 320): Int

tools/SourceKit/docs/Protocol.md

@@ -697,7 +697,46 @@ Welcome to SourceKit.  Type ':help' for assistance.
 }
 ```
 
+## Expression Type
+This request collects the types of all expressions in a source file after type checking.
+To fulfill this task, the client must provide the path to the Swift source file under
+type checking and the necessary compiler arguments to help resolve all dependencies.
 
+### Request
+
+```
+{
+    <key.request>:            (UID)     <source.request.expression.type>,
+    <key.sourcefile>:         (string)  // Absolute path to the file.
+    <key.compilerargs>:       [string*] // Array of zero or more strings for the compiler arguments,
+                                        // e.g ["-sdk", "/path/to/sdk"]. If key.sourcefile is provided,
+                                        // these must include the path to that file.
+}
+```
+
+### Response
+```
+{
+    <key.printedtypebuffer>:          (string)                    // A text buffer where all expression types are printed to.
+    <key.expression_type_list>:       (array) [expr-type-info*]   // A list of expression and type
+}
+```
+
+```
+expr-type-info ::=
+{
+  <key.expression_offset>:    (int64)    // Offset of an expression in the source file
+  <key.expression_length>:    (int64)    // Length of an expression in the source file
+  <key.type_offset>:          (int64)    // Offset of the printed type of the expression in the printed type buffer
+  <key.type_length>:          (int64)    // Length of the printed type of the expression in the printed type buffer
+}
+```
+
+### Testing
+
+```
+$ sourcekitd-test -req=collect-type /path/to/file.swift -- /path/to/file.swift
+```
 
 # UIDs
 

tools/SourceKit/include/SourceKit/Core/LangSupport.h

@@ -123,6 +123,17 @@ struct CodeCompletionInfo {
   Optional<ArrayRef<ParameterStructure>> parametersStructure;
 };
 
+struct ExpressionType {
+  unsigned ExprOffset;
+  unsigned ExprLength;
+  unsigned TypeOffset;
+};
+
+struct ExpressionTypesInFile {
+  std::vector<ExpressionType> Results;
+  StringRef TypeBuffer;
+};
+
 class CodeCompletionConsumer {
   virtual void anchor();
 
@@ -708,6 +719,10 @@ class LangSupport {
                                    ArrayRef<const char*> Args,
                                    CategorizedEditsReceiver Receiver) = 0;
 
+  virtual void collectExpressionTypes(StringRef FileName,
+                                      ArrayRef<const char *> Args,
+                                      std::function<void(const ExpressionTypesInFile&)> Receiver) = 0;
+
   virtual void getDocInfo(llvm::MemoryBuffer *InputBuf,
                           StringRef ModuleName,
                           ArrayRef<const char *> Args,
@@ -726,7 +741,6 @@ class LangSupport {
 
   virtual void getStatistics(StatisticsReceiver) = 0;
 };
-
 } // namespace SourceKit
 
 #endif