1//===--- extra/module-map-checker/CoverageChecker.cpp -------------------===//
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// This file implements a class that validates a module map by checking that
10// all headers in the corresponding directories are accounted for.
11//
12// This class uses a previously loaded module map object.
13// Starting at the module map file directory, or just the include
14// paths, if specified, it will collect the names of all the files it
15// considers headers (no extension, .h, or .inc--if you need more, modify the
16// ModularizeUtilities::isHeader function).
17// It then compares the headers against those referenced
18// in the module map, either explicitly named, or implicitly named via an
19// umbrella directory or umbrella file, as parsed by the ModuleMap object.
20// If headers are found which are not referenced or covered by an umbrella
21// directory or file, warning messages will be produced, and the doChecks
22// function will return an error code of 1. Other errors result in an error
23// code of 2. If no problems are found, an error code of 0 is returned.
24//
25// Note that in the case of umbrella headers, this tool invokes the compiler
26// to preprocess the file, and uses a callback to collect the header files
27// included by the umbrella header or any of its nested includes. If any
28// front end options are needed for these compiler invocations, these are
29// to be passed in via the CommandLine parameter.
30//
31// Warning message have the form:
32//
33// warning: module.modulemap does not account for file: Level3A.h
34//
35// Note that for the case of the module map referencing a file that does
36// not exist, the module map parser in Clang will (at the time of this
37// writing) display an error message.
38//
39// Potential problems with this program:
40//
41// 1. Might need a better header matching mechanism, or extensions to the
42// canonical file format used.
43//
44// 2. It might need to support additional header file extensions.
45//
46// Future directions:
47//
48// 1. Add an option to fix the problems found, writing a new module map.
49// Include an extra option to add unaccounted-for headers as excluded.
50//
51//===----------------------------------------------------------------------===//
52
53#include "ModularizeUtilities.h"
54#include "clang/AST/ASTConsumer.h"
55#include "CoverageChecker.h"
56#include "clang/AST/ASTContext.h"
57#include "clang/AST/RecursiveASTVisitor.h"
58#include "clang/Basic/SourceManager.h"
59#include "clang/Driver/Options.h"
60#include "clang/Frontend/CompilerInstance.h"
61#include "clang/Frontend/FrontendAction.h"
62#include "clang/Frontend/FrontendActions.h"
63#include "clang/Lex/PPCallbacks.h"
64#include "clang/Lex/Preprocessor.h"
65#include "clang/Tooling/CompilationDatabase.h"
66#include "clang/Tooling/Tooling.h"
67#include "llvm/Option/Option.h"
68#include "llvm/Support/CommandLine.h"
69#include "llvm/Support/FileSystem.h"
70#include "llvm/Support/Path.h"
71#include "llvm/Support/raw_ostream.h"
72
73using namespace Modularize;
74using namespace clang;
75using namespace clang::driver;
76using namespace clang::driver::options;
77using namespace clang::tooling;
78namespace cl = llvm::cl;
79namespace sys = llvm::sys;
80
81// Preprocessor callbacks.
82// We basically just collect include files.
83class CoverageCheckerCallbacks : public PPCallbacks {
84public:
85 CoverageCheckerCallbacks(CoverageChecker &Checker) : Checker(Checker) {}
86 ~CoverageCheckerCallbacks() override {}
87
88 // Include directive callback.
89 void InclusionDirective(SourceLocation HashLoc, const Token &IncludeTok,
90 StringRef FileName, bool IsAngled,
91 CharSourceRange FilenameRange,
92 OptionalFileEntryRef File, StringRef SearchPath,
93 StringRef RelativePath, const Module *Imported,
94 SrcMgr::CharacteristicKind FileType) override {
95 Checker.collectUmbrellaHeaderHeader(File->getName());
96 }
97
98private:
99 CoverageChecker &Checker;
100};
101
102// Frontend action stuff:
103
104// Consumer is responsible for setting up the callbacks.
105class CoverageCheckerConsumer : public ASTConsumer {
106public:
107 CoverageCheckerConsumer(CoverageChecker &Checker, Preprocessor &PP) {
108 // PP takes ownership.
109 PP.addPPCallbacks(std::make_unique<CoverageCheckerCallbacks>(Checker));
110 }
111};
112
113class CoverageCheckerAction : public SyntaxOnlyAction {
114public:
115 CoverageCheckerAction(CoverageChecker &Checker) : Checker(Checker) {}
116
117protected:
118 std::unique_ptr<ASTConsumer> CreateASTConsumer(CompilerInstance &CI,
119 StringRef InFile) override {
120 return std::make_unique<CoverageCheckerConsumer>(Checker,
121 CI.getPreprocessor());
122 }
123
124private:
125 CoverageChecker &Checker;
126};
127
128class CoverageCheckerFrontendActionFactory : public FrontendActionFactory {
129public:
130 CoverageCheckerFrontendActionFactory(CoverageChecker &Checker)
131 : Checker(Checker) {}
132
133 std::unique_ptr<FrontendAction> create() override {
134 return std::make_unique<CoverageCheckerAction>(Checker);
135 }
136
137private:
138 CoverageChecker &Checker;
139};
140
141// CoverageChecker class implementation.
142
143// Constructor.
144CoverageChecker::CoverageChecker(StringRef ModuleMapPath,
145 std::vector<std::string> &IncludePaths,
146 ArrayRef<std::string> CommandLine,
147 clang::ModuleMap *ModuleMap)
148 : ModuleMapPath(ModuleMapPath), IncludePaths(IncludePaths),
149 CommandLine(CommandLine),
150 ModMap(ModuleMap) {}
151
152// Create instance of CoverageChecker, to simplify setting up
153// subordinate objects.
154std::unique_ptr<CoverageChecker> CoverageChecker::createCoverageChecker(
155 StringRef ModuleMapPath, std::vector<std::string> &IncludePaths,
156 ArrayRef<std::string> CommandLine, clang::ModuleMap *ModuleMap) {
157
158 return std::make_unique<CoverageChecker>(ModuleMapPath, IncludePaths,
159 CommandLine, ModuleMap);
160}
161
162// Do checks.
163// Starting from the directory of the module.modulemap file,
164// Find all header files, optionally looking only at files
165// covered by the include path options, and compare against
166// the headers referenced by the module.modulemap file.
167// Display warnings for unaccounted-for header files.
168// Returns error_code of 0 if there were no errors or warnings, 1 if there
169// were warnings, 2 if any other problem, such as if a bad
170// module map path argument was specified.
171std::error_code CoverageChecker::doChecks() {
172 std::error_code returnValue;
173
174 // Collect the headers referenced in the modules.
175 collectModuleHeaders();
176
177 // Collect the file system headers.
178 if (!collectFileSystemHeaders())
179 return std::error_code(2, std::generic_category());
180
181 // Do the checks. These save the problematic file names.
182 findUnaccountedForHeaders();
183
184 // Check for warnings.
185 if (!UnaccountedForHeaders.empty())
186 returnValue = std::error_code(1, std::generic_category());
187
188 return returnValue;
189}
190
191// The following functions are called by doChecks.
192
193// Collect module headers.
194// Walks the modules and collects referenced headers into
195// ModuleMapHeadersSet.
196void CoverageChecker::collectModuleHeaders() {
197 for (ModuleMap::module_iterator I = ModMap->module_begin(),
198 E = ModMap->module_end();
199 I != E; ++I) {
200 collectModuleHeaders(*I->second);
201 }
202}
203
204// Collect referenced headers from one module.
205// Collects the headers referenced in the given module into
206// ModuleMapHeadersSet.
207// FIXME: Doesn't collect files from umbrella header.
208bool CoverageChecker::collectModuleHeaders(const Module &Mod) {
209
210 if (std::optional<Module::Header> UmbrellaHeader =
211 Mod.getUmbrellaHeaderAsWritten()) {
212 // Collect umbrella header.
213 ModuleMapHeadersSet.insert(
214 ModularizeUtilities::getCanonicalPath(UmbrellaHeader->Entry.getName()));
215 // Preprocess umbrella header and collect the headers it references.
216 if (!collectUmbrellaHeaderHeaders(UmbrellaHeader->Entry.getName()))
217 return false;
218 } else if (std::optional<Module::DirectoryName> UmbrellaDir =
219 Mod.getUmbrellaDirAsWritten()) {
220 // Collect headers in umbrella directory.
221 if (!collectUmbrellaHeaders(UmbrellaDir->Entry.getName()))
222 return false;
223 }
224
225 for (auto &HeaderKind : Mod.Headers)
226 for (auto &Header : HeaderKind)
227 ModuleMapHeadersSet.insert(
228 ModularizeUtilities::getCanonicalPath(Header.Entry.getName()));
229
230 for (auto *Submodule : Mod.submodules())
231 collectModuleHeaders(*Submodule);
232
233 return true;
234}
235
236// Collect headers from an umbrella directory.
237bool CoverageChecker::collectUmbrellaHeaders(StringRef UmbrellaDirName) {
238 // Initialize directory name.
239 SmallString<256> Directory(ModuleMapDirectory);
240 if (UmbrellaDirName.size())
241 sys::path::append(Directory, UmbrellaDirName);
242 if (Directory.size() == 0)
243 Directory = ".";
244 // Walk the directory.
245 std::error_code EC;
246 for (sys::fs::directory_iterator I(Directory.str(), EC), E; I != E;
247 I.increment(EC)) {
248 if (EC)
249 return false;
250 std::string File(I->path());
251 llvm::ErrorOr<sys::fs::basic_file_status> Status = I->status();
252 if (!Status)
253 return false;
254 sys::fs::file_type Type = Status->type();
255 // If the file is a directory, ignore the name and recurse.
256 if (Type == sys::fs::file_type::directory_file) {
257 if (!collectUmbrellaHeaders(File))
258 return false;
259 continue;
260 }
261 // If the file does not have a common header extension, ignore it.
262 if (!ModularizeUtilities::isHeader(File))
263 continue;
264 // Save header name.
265 ModuleMapHeadersSet.insert(ModularizeUtilities::getCanonicalPath(File));
266 }
267 return true;
268}
269
270// Collect headers referenced from an umbrella file.
271bool
272CoverageChecker::collectUmbrellaHeaderHeaders(StringRef UmbrellaHeaderName) {
273
274 SmallString<256> PathBuf(ModuleMapDirectory);
275
276 // If directory is empty, it's the current directory.
277 if (ModuleMapDirectory.length() == 0)
278 sys::fs::current_path(PathBuf);
279
280 // Create the compilation database.
281 std::unique_ptr<CompilationDatabase> Compilations;
282 Compilations.reset(new FixedCompilationDatabase(Twine(PathBuf), CommandLine));
283
284 std::vector<std::string> HeaderPath;
285 HeaderPath.push_back(std::string(UmbrellaHeaderName));
286
287 // Create the tool and run the compilation.
288 ClangTool Tool(*Compilations, HeaderPath);
289 int HadErrors = Tool.run(new CoverageCheckerFrontendActionFactory(*this));
290
291 // If we had errors, exit early.
292 return !HadErrors;
293}
294
295// Called from CoverageCheckerCallbacks to track a header included
296// from an umbrella header.
297void CoverageChecker::collectUmbrellaHeaderHeader(StringRef HeaderName) {
298
299 SmallString<256> PathBuf(ModuleMapDirectory);
300 // If directory is empty, it's the current directory.
301 if (ModuleMapDirectory.length() == 0)
302 sys::fs::current_path(PathBuf);
303 // HeaderName will have an absolute path, so if it's the module map
304 // directory, we remove it, also skipping trailing separator.
305 if (HeaderName.startswith(PathBuf))
306 HeaderName = HeaderName.substr(PathBuf.size() + 1);
307 // Save header name.
308 ModuleMapHeadersSet.insert(ModularizeUtilities::getCanonicalPath(HeaderName));
309}
310
311// Collect file system header files.
312// This function scans the file system for header files,
313// starting at the directory of the module.modulemap file,
314// optionally filtering out all but the files covered by
315// the include path options.
316// Returns true if no errors.
317bool CoverageChecker::collectFileSystemHeaders() {
318
319 // Get directory containing the module.modulemap file.
320 // Might be relative to current directory, absolute, or empty.
321 ModuleMapDirectory = ModularizeUtilities::getDirectoryFromPath(ModuleMapPath);
322
323 // If no include paths specified, we do the whole tree starting
324 // at the module.modulemap directory.
325 if (IncludePaths.size() == 0) {
326 if (!collectFileSystemHeaders(StringRef("")))
327 return false;
328 }
329 else {
330 // Otherwise we only look at the sub-trees specified by the
331 // include paths.
332 for (std::vector<std::string>::const_iterator I = IncludePaths.begin(),
333 E = IncludePaths.end();
334 I != E; ++I) {
335 if (!collectFileSystemHeaders(*I))
336 return false;
337 }
338 }
339
340 // Sort it, because different file systems might order the file differently.
341 llvm::sort(FileSystemHeaders);
342
343 return true;
344}
345
346// Collect file system header files from the given path.
347// This function scans the file system for header files,
348// starting at the given directory, which is assumed to be
349// relative to the directory of the module.modulemap file.
350// \returns True if no errors.
351bool CoverageChecker::collectFileSystemHeaders(StringRef IncludePath) {
352
353 // Initialize directory name.
354 SmallString<256> Directory(ModuleMapDirectory);
355 if (IncludePath.size())
356 sys::path::append(Directory, IncludePath);
357 if (Directory.size() == 0)
358 Directory = ".";
359 if (IncludePath.startswith("/") || IncludePath.startswith("\\") ||
360 ((IncludePath.size() >= 2) && (IncludePath[1] == ':'))) {
361 llvm::errs() << "error: Include path \"" << IncludePath
362 << "\" is not relative to the module map file.\n";
363 return false;
364 }
365
366 // Recursively walk the directory tree.
367 std::error_code EC;
368 int Count = 0;
369 for (sys::fs::recursive_directory_iterator I(Directory.str(), EC), E; I != E;
370 I.increment(EC)) {
371 if (EC)
372 return false;
373 //std::string file(I->path());
374 StringRef file(I->path());
375 llvm::ErrorOr<sys::fs::basic_file_status> Status = I->status();
376 if (!Status)
377 return false;
378 sys::fs::file_type type = Status->type();
379 // If the file is a directory, ignore the name (but still recurses).
380 if (type == sys::fs::file_type::directory_file)
381 continue;
382 // Assume directories or files starting with '.' are private and not to
383 // be considered.
384 if (file.contains("\\.") || file.contains("/."))
385 continue;
386 // If the file does not have a common header extension, ignore it.
387 if (!ModularizeUtilities::isHeader(file))
388 continue;
389 // Save header name.
390 FileSystemHeaders.push_back(ModularizeUtilities::getCanonicalPath(file));
391 Count++;
392 }
393 if (Count == 0) {
394 llvm::errs() << "warning: No headers found in include path: \""
395 << IncludePath << "\"\n";
396 }
397 return true;
398}
399
400// Find headers unaccounted-for in module map.
401// This function compares the list of collected header files
402// against those referenced in the module map. Display
403// warnings for unaccounted-for header files.
404// Save unaccounted-for file list for possible.
405// fixing action.
406// FIXME: There probably needs to be some canonalization
407// of file names so that header path can be correctly
408// matched. Also, a map could be used for the headers
409// referenced in the module, but
410void CoverageChecker::findUnaccountedForHeaders() {
411 // Walk over file system headers.
412 for (std::vector<std::string>::const_iterator I = FileSystemHeaders.begin(),
413 E = FileSystemHeaders.end();
414 I != E; ++I) {
415 // Look for header in module map.
416 if (ModuleMapHeadersSet.insert(*I).second) {
417 UnaccountedForHeaders.push_back(*I);
418 llvm::errs() << "warning: " << ModuleMapPath
419 << " does not account for file: " << *I << "\n";
420 }
421 }
422}
423