1 //===- CompilationDatabase.h ------------------------------------*- C++ -*-===//
2 //
3 //                     The LLVM Compiler Infrastructure
4 //
5 // This file is distributed under the University of Illinois Open Source
6 // License. See LICENSE.TXT for details.
7 //
8 //===----------------------------------------------------------------------===//
9 //
10 //  This file provides an interface and multiple implementations for
11 //  CompilationDatabases.
12 //
13 //  While C++ refactoring and analysis tools are not compilers, and thus
14 //  don't run as part of the build system, they need the exact information
15 //  of a build in order to be able to correctly understand the C++ code of
16 //  the project. This information is provided via the CompilationDatabase
17 //  interface.
18 //
19 //  To create a CompilationDatabase from a build directory one can call
20 //  CompilationDatabase::loadFromDirectory(), which deduces the correct
21 //  compilation database from the root of the build tree.
22 //
23 //  See the concrete subclasses of CompilationDatabase for currently supported
24 //  formats.
25 //
26 //===----------------------------------------------------------------------===//
27 
28 #ifndef LLVM_CLANG_TOOLING_COMPILATIONDATABASE_H
29 #define LLVM_CLANG_TOOLING_COMPILATIONDATABASE_H
30 
31 #include "clang/Basic/LLVM.h"
32 #include "llvm/ADT/ArrayRef.h"
33 #include "llvm/ADT/StringRef.h"
34 #include "llvm/ADT/Twine.h"
35 #include <memory>
36 #include <string>
37 #include <utility>
38 #include <vector>
39 
40 namespace clang {
41 namespace tooling {
42 
43 /// Specifies the working directory and command of a compilation.
44 struct CompileCommand {
45   CompileCommand() = default;
46   CompileCommand(Twine Directory, Twine Filename,
47                  std::vector<std::string> CommandLine, Twine Output)
48       : Directory(Directory.str()), Filename(Filename.str()),
49         CommandLine(std::move(CommandLine)), Output(Output.str()){}
50 
51   /// The working directory the command was executed from.
52   std::string Directory;
53 
54   /// The source file associated with the command.
55   std::string Filename;
56 
57   /// The command line that was executed.
58   std::vector<std::string> CommandLine;
59 
60   /// The output file associated with the command.
61   std::string Output;
62 
63   friend bool operator==(const CompileCommand &LHS, const CompileCommand &RHS) {
64     return LHS.Directory == RHS.Directory && LHS.Filename == RHS.Filename &&
65            LHS.CommandLine == RHS.CommandLine && LHS.Output == RHS.Output;
66   }
67 
68   friend bool operator!=(const CompileCommand &LHS, const CompileCommand &RHS) {
69     return !(LHS == RHS);
70   }
71 };
72 
73 /// Interface for compilation databases.
74 ///
75 /// A compilation database allows the user to retrieve compile command lines
76 /// for the files in a project.
77 ///
78 /// Many implementations are enumerable, allowing all command lines to be
79 /// retrieved. These can be used to run clang tools over a subset of the files
80 /// in a project.
81 class CompilationDatabase {
82 public:
83   virtual ~CompilationDatabase();
84 
85   /// Loads a compilation database from a build directory.
86   ///
87   /// Looks at the specified 'BuildDirectory' and creates a compilation database
88   /// that allows to query compile commands for source files in the
89   /// corresponding source tree.
90   ///
91   /// Returns NULL and sets ErrorMessage if we were not able to build up a
92   /// compilation database for the build directory.
93   ///
94   /// FIXME: Currently only supports JSON compilation databases, which
95   /// are named 'compile_commands.json' in the given directory. Extend this
96   /// for other build types (like ninja build files).
97   static std::unique_ptr<CompilationDatabase>
98   loadFromDirectory(StringRef BuildDirectory, std::string &ErrorMessage);
99 
100   /// Tries to detect a compilation database location and load it.
101   ///
102   /// Looks for a compilation database in all parent paths of file 'SourceFile'
103   /// by calling loadFromDirectory.
104   static std::unique_ptr<CompilationDatabase>
105   autoDetectFromSource(StringRef SourceFile, std::string &ErrorMessage);
106 
107   /// Tries to detect a compilation database location and load it.
108   ///
109   /// Looks for a compilation database in directory 'SourceDir' and all
110   /// its parent paths by calling loadFromDirectory.
111   static std::unique_ptr<CompilationDatabase>
112   autoDetectFromDirectory(StringRef SourceDir, std::string &ErrorMessage);
113 
114   /// Returns all compile commands in which the specified file was
115   /// compiled.
116   ///
117   /// This includes compile commands that span multiple source files.
118   /// For example, consider a project with the following compilations:
119   /// $ clang++ -o test a.cc b.cc t.cc
120   /// $ clang++ -o production a.cc b.cc -DPRODUCTION
121   /// A compilation database representing the project would return both command
122   /// lines for a.cc and b.cc and only the first command line for t.cc.
123   virtual std::vector<CompileCommand> getCompileCommands(
124       StringRef FilePath) const = 0;
125 
126   /// Returns the list of all files available in the compilation database.
127   ///
128   /// By default, returns nothing. Implementations should override this if they
129   /// can enumerate their source files.
130   virtual std::vector<std::string> getAllFiles() const { return {}; }
131 
132   /// Returns all compile commands for all the files in the compilation
133   /// database.
134   ///
135   /// FIXME: Add a layer in Tooling that provides an interface to run a tool
136   /// over all files in a compilation database. Not all build systems have the
137   /// ability to provide a feasible implementation for \c getAllCompileCommands.
138   ///
139   /// By default, this is implemented in terms of getAllFiles() and
140   /// getCompileCommands(). Subclasses may override this for efficiency.
141   virtual std::vector<CompileCommand> getAllCompileCommands() const;
142 };
143 
144 /// A compilation database that returns a single compile command line.
145 ///
146 /// Useful when we want a tool to behave more like a compiler invocation.
147 /// This compilation database is not enumerable: getAllFiles() returns {}.
148 class FixedCompilationDatabase : public CompilationDatabase {
149 public:
150   /// Creates a FixedCompilationDatabase from the arguments after "--".
151   ///
152   /// Parses the given command line for "--". If "--" is found, the rest of
153   /// the arguments will make up the command line in the returned
154   /// FixedCompilationDatabase.
155   /// The arguments after "--" must not include positional parameters or the
156   /// argv[0] of the tool. Those will be added by the FixedCompilationDatabase
157   /// when a CompileCommand is requested. The argv[0] of the returned command
158   /// line will be "clang-tool".
159   ///
160   /// Returns NULL in case "--" is not found.
161   ///
162   /// The argument list is meant to be compatible with normal llvm command line
163   /// parsing in main methods.
164   /// int main(int argc, char **argv) {
165   ///   std::unique_ptr<FixedCompilationDatabase> Compilations(
166   ///     FixedCompilationDatabase::loadFromCommandLine(argc, argv));
167   ///   cl::ParseCommandLineOptions(argc, argv);
168   ///   ...
169   /// }
170   ///
171   /// \param Argc The number of command line arguments - will be changed to
172   /// the number of arguments before "--", if "--" was found in the argument
173   /// list.
174   /// \param Argv Points to the command line arguments.
175   /// \param ErrorMsg Contains error text if the function returns null pointer.
176   /// \param Directory The base directory used in the FixedCompilationDatabase.
177   static std::unique_ptr<FixedCompilationDatabase> loadFromCommandLine(
178       int &Argc, const char *const *Argv, std::string &ErrorMsg,
179       Twine Directory = ".");
180 
181   /// Reads flags from the given file, one-per line.
182   /// Returns nullptr and sets ErrorMessage if we can't read the file.
183   static std::unique_ptr<FixedCompilationDatabase>
184   loadFromFile(StringRef Path, std::string &ErrorMsg);
185 
186   /// Constructs a compilation data base from a specified directory
187   /// and command line.
188   FixedCompilationDatabase(Twine Directory, ArrayRef<std::string> CommandLine);
189 
190   /// Returns the given compile command.
191   ///
192   /// Will always return a vector with one entry that contains the directory
193   /// and command line specified at construction with "clang-tool" as argv[0]
194   /// and 'FilePath' as positional argument.
195   std::vector<CompileCommand>
196   getCompileCommands(StringRef FilePath) const override;
197 
198 private:
199   /// This is built up to contain a single entry vector to be returned from
200   /// getCompileCommands after adding the positional argument.
201   std::vector<CompileCommand> CompileCommands;
202 };
203 
204 /// Returns a wrapped CompilationDatabase that defers to the provided one,
205 /// but getCompileCommands() will infer commands for unknown files.
206 /// The return value of getAllFiles() or getAllCompileCommands() is unchanged.
207 /// See InterpolatingCompilationDatabase.cpp for details on heuristics.
208 std::unique_ptr<CompilationDatabase>
209     inferMissingCompileCommands(std::unique_ptr<CompilationDatabase>);
210 
211 } // namespace tooling
212 } // namespace clang
213 
214 #endif // LLVM_CLANG_TOOLING_COMPILATIONDATABASE_H
215