1//===--- FileManager.h - File System Probing and Caching --------*- C++ -*-===//
2//
3// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
4// See https://llvm.org/LICENSE.txt for license information.
5// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
6//
7//===----------------------------------------------------------------------===//
8///
9/// \file
10/// Defines the clang::FileManager interface and associated types.
11///
12//===----------------------------------------------------------------------===//
13
14#ifndef LLVM_CLANG_BASIC_FILEMANAGER_H
15#define LLVM_CLANG_BASIC_FILEMANAGER_H
16
17#include "clang/Basic/DirectoryEntry.h"
18#include "clang/Basic/FileEntry.h"
19#include "clang/Basic/FileSystemOptions.h"
20#include "clang/Basic/LLVM.h"
21#include "llvm/ADT/DenseMap.h"
22#include "llvm/ADT/IntrusiveRefCntPtr.h"
23#include "llvm/ADT/PointerUnion.h"
24#include "llvm/ADT/SmallVector.h"
25#include "llvm/ADT/StringMap.h"
26#include "llvm/ADT/StringRef.h"
27#include "llvm/Support/Allocator.h"
28#include "llvm/Support/ErrorOr.h"
29#include "llvm/Support/FileSystem.h"
30#include "llvm/Support/VirtualFileSystem.h"
31#include <ctime>
32#include <map>
33#include <memory>
34#include <string>
35
36namespace llvm {
37
38class MemoryBuffer;
39
40} // end namespace llvm
41
42namespace clang {
43
44class FileSystemStatCache;
45
46/// Implements support for file system lookup, file system caching,
47/// and directory search management.
48///
49/// This also handles more advanced properties, such as uniquing files based
50/// on "inode", so that a file with two names (e.g. symlinked) will be treated
51/// as a single file.
52///
53class FileManager : public RefCountedBase<FileManager> {
54 IntrusiveRefCntPtr<llvm::vfs::FileSystem> FS;
55 FileSystemOptions FileSystemOpts;
56 llvm::SpecificBumpPtrAllocator<FileEntry> FilesAlloc;
57 llvm::SpecificBumpPtrAllocator<DirectoryEntry> DirsAlloc;
58
59 /// Cache for existing real directories.
60 llvm::DenseMap<llvm::sys::fs::UniqueID, DirectoryEntry *> UniqueRealDirs;
61
62 /// Cache for existing real files.
63 llvm::DenseMap<llvm::sys::fs::UniqueID, FileEntry *> UniqueRealFiles;
64
65 /// The virtual directories that we have allocated.
66 ///
67 /// For each virtual file (e.g. foo/bar/baz.cpp), we add all of its parent
68 /// directories (foo/ and foo/bar/) here.
69 SmallVector<DirectoryEntry *, 4> VirtualDirectoryEntries;
70 /// The virtual files that we have allocated.
71 SmallVector<FileEntry *, 4> VirtualFileEntries;
72
73 /// A set of files that bypass the maps and uniquing. They can have
74 /// conflicting filenames.
75 SmallVector<FileEntry *, 0> BypassFileEntries;
76
77 /// A cache that maps paths to directory entries (either real or
78 /// virtual) we have looked up, or an error that occurred when we looked up
79 /// the directory.
80 ///
81 /// The actual Entries for real directories/files are
82 /// owned by UniqueRealDirs/UniqueRealFiles above, while the Entries
83 /// for virtual directories/files are owned by
84 /// VirtualDirectoryEntries/VirtualFileEntries above.
85 ///
86 llvm::StringMap<llvm::ErrorOr<DirectoryEntry &>, llvm::BumpPtrAllocator>
87 SeenDirEntries;
88
89 /// A cache that maps paths to file entries (either real or
90 /// virtual) we have looked up, or an error that occurred when we looked up
91 /// the file.
92 ///
93 /// \see SeenDirEntries
94 llvm::StringMap<llvm::ErrorOr<FileEntryRef::MapValue>, llvm::BumpPtrAllocator>
95 SeenFileEntries;
96
97 /// A mirror of SeenFileEntries to give fake answers for getBypassFile().
98 ///
99 /// Don't bother hooking up a BumpPtrAllocator. This should be rarely used,
100 /// and only on error paths.
101 std::unique_ptr<llvm::StringMap<llvm::ErrorOr<FileEntryRef::MapValue>>>
102 SeenBypassFileEntries;
103
104 /// The file entry for stdin, if it has been accessed through the FileManager.
105 OptionalFileEntryRef STDIN;
106
107 /// The canonical names of files and directories .
108 llvm::DenseMap<const void *, llvm::StringRef> CanonicalNames;
109
110 /// Storage for canonical names that we have computed.
111 llvm::BumpPtrAllocator CanonicalNameStorage;
112
113 /// Each FileEntry we create is assigned a unique ID #.
114 ///
115 unsigned NextFileUID;
116
117 /// Statistics gathered during the lifetime of the FileManager.
118 unsigned NumDirLookups = 0;
119 unsigned NumFileLookups = 0;
120 unsigned NumDirCacheMisses = 0;
121 unsigned NumFileCacheMisses = 0;
122
123 // Caching.
124 std::unique_ptr<FileSystemStatCache> StatCache;
125
126 std::error_code getStatValue(StringRef Path, llvm::vfs::Status &Status,
127 bool isFile, std::unique_ptr<llvm::vfs::File> *F,
128 bool IsText = true);
129
130 /// Add all ancestors of the given path (pointing to either a file
131 /// or a directory) as virtual directories.
132 void addAncestorsAsVirtualDirs(StringRef Path);
133
134 /// Fills the RealPathName in file entry.
135 void fillRealPathName(FileEntry *UFE, llvm::StringRef FileName);
136
137public:
138 /// Construct a file manager, optionally with a custom VFS.
139 ///
140 /// \param FS if non-null, the VFS to use. Otherwise uses
141 /// llvm::vfs::getRealFileSystem().
142 FileManager(const FileSystemOptions &FileSystemOpts,
143 IntrusiveRefCntPtr<llvm::vfs::FileSystem> FS = nullptr);
144 ~FileManager();
145
146 /// Installs the provided FileSystemStatCache object within
147 /// the FileManager.
148 ///
149 /// Ownership of this object is transferred to the FileManager.
150 ///
151 /// \param statCache the new stat cache to install. Ownership of this
152 /// object is transferred to the FileManager.
153 void setStatCache(std::unique_ptr<FileSystemStatCache> statCache);
154
155 /// Removes the FileSystemStatCache object from the manager.
156 void clearStatCache();
157
158 /// Returns the number of unique real file entries cached by the file manager.
159 size_t getNumUniqueRealFiles() const { return UniqueRealFiles.size(); }
160
161 /// Lookup, cache, and verify the specified directory (real or
162 /// virtual).
163 ///
164 /// This returns a \c std::error_code if there was an error reading the
165 /// directory. On success, returns the reference to the directory entry
166 /// together with the exact path that was used to access a file by a
167 /// particular call to getDirectoryRef.
168 ///
169 /// \param CacheFailure If true and the file does not exist, we'll cache
170 /// the failure to find this file.
171 llvm::Expected<DirectoryEntryRef> getDirectoryRef(StringRef DirName,
172 bool CacheFailure = true);
173
174 /// Get a \c DirectoryEntryRef if it exists, without doing anything on error.
175 OptionalDirectoryEntryRef getOptionalDirectoryRef(StringRef DirName,
176 bool CacheFailure = true) {
177 return llvm::expectedToOptional(E: getDirectoryRef(DirName, CacheFailure));
178 }
179
180 /// Lookup, cache, and verify the specified file (real or virtual). Return the
181 /// reference to the file entry together with the exact path that was used to
182 /// access a file by a particular call to getFileRef. If the underlying VFS is
183 /// a redirecting VFS that uses external file names, the returned FileEntryRef
184 /// will use the external name instead of the filename that was passed to this
185 /// method.
186 ///
187 /// This returns a \c std::error_code if there was an error loading the file,
188 /// or a \c FileEntryRef otherwise.
189 ///
190 /// \param OpenFile if true and the file exists, it will be opened.
191 ///
192 /// \param CacheFailure If true and the file does not exist, we'll cache
193 /// the failure to find this file.
194 llvm::Expected<FileEntryRef> getFileRef(StringRef Filename,
195 bool OpenFile = false,
196 bool CacheFailure = true,
197 bool IsText = true);
198
199 /// Get the FileEntryRef for stdin, returning an error if stdin cannot be
200 /// read.
201 ///
202 /// This reads and caches stdin before returning. Subsequent calls return the
203 /// same file entry, and a reference to the cached input is returned by calls
204 /// to getBufferForFile.
205 llvm::Expected<FileEntryRef> getSTDIN();
206
207 /// Get a FileEntryRef if it exists, without doing anything on error.
208 OptionalFileEntryRef getOptionalFileRef(StringRef Filename,
209 bool OpenFile = false,
210 bool CacheFailure = true) {
211 return llvm::expectedToOptional(
212 E: getFileRef(Filename, OpenFile, CacheFailure));
213 }
214
215 /// Returns the current file system options
216 FileSystemOptions &getFileSystemOpts() { return FileSystemOpts; }
217 const FileSystemOptions &getFileSystemOpts() const { return FileSystemOpts; }
218
219 llvm::vfs::FileSystem &getVirtualFileSystem() const { return *FS; }
220 llvm::IntrusiveRefCntPtr<llvm::vfs::FileSystem>
221 getVirtualFileSystemPtr() const {
222 return FS;
223 }
224
225 /// Enable or disable tracking of VFS usage. Used to not track full header
226 /// search and implicit modulemap lookup.
227 void trackVFSUsage(bool Active);
228
229 void setVirtualFileSystem(IntrusiveRefCntPtr<llvm::vfs::FileSystem> FS) {
230 this->FS = std::move(FS);
231 }
232
233 /// Retrieve a file entry for a "virtual" file that acts as
234 /// if there were a file with the given name on disk.
235 ///
236 /// The file itself is not accessed.
237 FileEntryRef getVirtualFileRef(StringRef Filename, off_t Size,
238 time_t ModificationTime);
239
240 LLVM_DEPRECATED("Functions returning FileEntry are deprecated.",
241 "getVirtualFileRef()")
242 const FileEntry *getVirtualFile(StringRef Filename, off_t Size,
243 time_t ModificationTime);
244
245 /// Retrieve a FileEntry that bypasses VFE, which is expected to be a virtual
246 /// file entry, to access the real file. The returned FileEntry will have
247 /// the same filename as FE but a different identity and its own stat.
248 ///
249 /// This should be used only for rare error recovery paths because it
250 /// bypasses all mapping and uniquing, blindly creating a new FileEntry.
251 /// There is no attempt to deduplicate these; if you bypass the same file
252 /// twice, you get two new file entries.
253 OptionalFileEntryRef getBypassFile(FileEntryRef VFE);
254
255 /// Open the specified file as a MemoryBuffer, returning a new
256 /// MemoryBuffer if successful, otherwise returning null.
257 /// The IsText parameter controls whether the file should be opened as a text
258 /// or binary file, and should be set to false if the file contents should be
259 /// treated as binary.
260 llvm::ErrorOr<std::unique_ptr<llvm::MemoryBuffer>>
261 getBufferForFile(FileEntryRef Entry, bool isVolatile = false,
262 bool RequiresNullTerminator = true,
263 std::optional<int64_t> MaybeLimit = std::nullopt,
264 bool IsText = true);
265 llvm::ErrorOr<std::unique_ptr<llvm::MemoryBuffer>>
266 getBufferForFile(StringRef Filename, bool isVolatile = false,
267 bool RequiresNullTerminator = true,
268 std::optional<int64_t> MaybeLimit = std::nullopt,
269 bool IsText = true) const {
270 return getBufferForFileImpl(Filename,
271 /*FileSize=*/FileSize: MaybeLimit.value_or(u: -1),
272 isVolatile, RequiresNullTerminator, IsText);
273 }
274
275private:
276 llvm::ErrorOr<std::unique_ptr<llvm::MemoryBuffer>>
277 getBufferForFileImpl(StringRef Filename, int64_t FileSize, bool isVolatile,
278 bool RequiresNullTerminator, bool IsText) const;
279
280 DirectoryEntry *&getRealDirEntry(const llvm::vfs::Status &Status);
281
282public:
283 /// Get the 'stat' information for the given \p Path.
284 ///
285 /// If the path is relative, it will be resolved against the WorkingDir of the
286 /// FileManager's FileSystemOptions.
287 ///
288 /// \returns a \c std::error_code describing an error, if there was one
289 std::error_code getNoncachedStatValue(StringRef Path,
290 llvm::vfs::Status &Result);
291
292 /// If path is not absolute and FileSystemOptions set the working
293 /// directory, the path is modified to be relative to the given
294 /// working directory.
295 /// \returns true if \c path changed.
296 bool FixupRelativePath(SmallVectorImpl<char> &path) const;
297
298 /// Makes \c Path absolute taking into account FileSystemOptions and the
299 /// working directory option.
300 /// \returns true if \c Path changed to absolute.
301 bool makeAbsolutePath(SmallVectorImpl<char> &Path) const;
302
303 /// Produce an array mapping from the unique IDs assigned to each
304 /// file to the corresponding FileEntryRef.
305 void
306 GetUniqueIDMapping(SmallVectorImpl<OptionalFileEntryRef> &UIDToFiles) const;
307
308 /// Retrieve the canonical name for a given directory.
309 ///
310 /// This is a very expensive operation, despite its results being cached,
311 /// and should only be used when the physical layout of the file system is
312 /// required, which is (almost) never.
313 StringRef getCanonicalName(DirectoryEntryRef Dir);
314
315 /// Retrieve the canonical name for a given file.
316 ///
317 /// This is a very expensive operation, despite its results being cached,
318 /// and should only be used when the physical layout of the file system is
319 /// required, which is (almost) never.
320 StringRef getCanonicalName(FileEntryRef File);
321
322private:
323 /// Retrieve the canonical name for a given file or directory.
324 ///
325 /// The first param is a key in the CanonicalNames array.
326 StringRef getCanonicalName(const void *Entry, StringRef Name);
327
328public:
329 void PrintStats() const;
330
331 /// Import statistics from a child FileManager and add them to this current
332 /// FileManager.
333 void AddStats(const FileManager &Other);
334};
335
336} // end namespace clang
337
338#endif // LLVM_CLANG_BASIC_FILEMANAGER_H
339