refs #341 - added fuzzing client

Makefile

@@ -4,22 +4,32 @@ CXXFLAGS = -Wall -Wextra -pedantic -Wcast-qual -Wfloat-equal -Wmissing-declarati
 LDFLAGS = -g $(LDOPTS)
 
 %.o: %.cpp	simplecpp.h
-	$(CXX) $(CXXFLAGS) -c $<
+	$(CXX) $(CXXFLAGS) -c $< $(LIB_FUZZING_ENGINE)
 
+fuzz_no.o: fuzz.cpp
+	$(CXX) $(CXXFLAGS) -DNO_FUZZ -c -o $@ fuzz.cpp
 
 testrunner:	test.o	simplecpp.o
-	$(CXX) $(LDFLAGS) simplecpp.o test.o -o testrunner
+	$(CXX) $(LDFLAGS) -o $@ $^
 
 test:	testrunner	simplecpp
 	./testrunner
 	python3 run-tests.py
 	python3 -m pytest integration_test.py -vv
 
+fuzz:	fuzz.o simplecpp.o
+	# TODO: use -stdlib=libc++ -lc++
+	# make fuzz CXX=clang++ CXXOPTS="-O2 -fno-omit-frame-pointer -g -gline-tables-only -DFUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION -fsanitize=address,undefined -fsanitize-address-use-after-scope -fno-sanitize=integer -fno-sanitize-recover=undefined" LIB_FUZZING_ENGINE="-fsanitize=fuzzer"
+	$(CXX) $(LDFLAGS) $(CXXFLAGS) -o $@ $^ $(LIB_FUZZING_ENGINE)
+
+no-fuzz:	fuzz_no.o simplecpp.o
+	$(CXX) $(LDFLAGS) $(CXXFLAGS) -o $@ $^
+
 selfcheck:	simplecpp
 	./selfcheck.sh
 
 simplecpp:	main.o simplecpp.o
-	$(CXX) $(LDFLAGS) main.o simplecpp.o -o simplecpp
+	$(CXX) $(LDFLAGS) -o $@ $^
 
 clean:
-	rm -f testrunner simplecpp *.o
+	rm -f testrunner fuzz no-fuzz simplecpp *.o

fuzz.cpp

@@ -0,0 +1,67 @@
+/*
+ * simplecpp - A simple and high-fidelity C/C++ preprocessor library
+ * Copyright (C) 2016-2024 simplecpp team
+ */
+
+#include "simplecpp.h"
+
+#include <cstdint>
+
+#ifdef NO_FUZZ
+#include <cstdlib>
+#include <fstream>
+#include <sstream>
+#include <string>
+#endif
+
+static void doProcess(const uint8_t *data, size_t dataSize)
+{
+    simplecpp::OutputList outputList;
+    std::vector<std::string> files;
+    simplecpp::TokenList rawtokens(data, dataSize, files, "test.cpp", &outputList);
+    rawtokens.removeComments();
+
+    simplecpp::TokenList outputTokens(files);
+    simplecpp::FileDataCache filedata;
+    simplecpp::DUI dui;
+    dui.removeComments = true;
+    std::list<simplecpp::MacroUsage> macroUsage;
+    std::list<simplecpp::IfCond> ifCond;
+    simplecpp::preprocess(outputTokens, rawtokens, files, filedata, dui, &outputList, &macroUsage, &ifCond);
+
+    simplecpp::cleanup(filedata);
+}
+
+#ifndef NO_FUZZ
+extern "C" int LLVMFuzzerTestOneInput(const uint8_t *data, size_t dataSize);
+
+int LLVMFuzzerTestOneInput(const uint8_t *data, size_t dataSize)
+{
+    doProcess(data, dataSize);
+    return 0;
+}
+#else
+int main(int argc, char * argv[])
+{
+    if (argc < 2 || argc > 3)
+        return EXIT_FAILURE;
+
+    std::ifstream f(argv[1]);
+    if (!f.is_open())
+        return EXIT_FAILURE;
+
+    std::ostringstream oss;
+    oss << f.rdbuf();
+
+    if (!f.good())
+        return EXIT_FAILURE;
+
+    const int cnt = (argc == 3) ? std::stoi(argv[2]) : 1;
+
+    const std::string code = oss.str();
+    for (int i = 0; i < cnt; ++i)
+        doProcess(reinterpret_cast<const uint8_t*>(code.data()), code.size());
+
+    return EXIT_SUCCESS;
+}
+#endif

main.cpp

@@ -121,6 +121,10 @@ int main(int argc, char **argv)
             std::cout << "error: could not open file '" << filename << "'" << std::endl;
             std::exit(1);
         }
+        if (!simplecpp::isFile(filename)) {
+            std::cout << "error: could not open file '" << filename << "' - not a regular file" << std::endl;
+            std::exit(1);
+        }
         rawtokens = new simplecpp::TokenList(f, files,filename,&outputList);
     } else {
         rawtokens = new simplecpp::TokenList(filename,files,&outputList);

simplecpp.cpp

@@ -42,8 +42,10 @@
 
 #ifdef _WIN32
 #  include <direct.h>
+using mode_t = unsigned short;
 #else
 #  include <sys/stat.h>
+#  include <sys/types.h>
 #endif
 
 static bool isHex(const std::string &s)
@@ -685,22 +687,35 @@ void simplecpp::TokenList::readfile(Stream &stream, const std::string &filename,
 
             if (oldLastToken != cback()) {
                 oldLastToken = cback();
-                if (!isLastLinePreprocessor())
+                const Token * const llTok = isLastLinePreprocessor();
+                if (!llTok)
                     continue;
-                const std::string lastline(lastLine());
-                if (lastline == "# file %str%") {
+                const Token * const llNextToken = llTok->next;
+                if (!llTok->next)
+                    continue;
+                // #file "file.c"
+                if (llNextToken->str() == "file" &&
+                    llNextToken->next &&
+                    llNextToken->next->str()[0] == '\"')
+                {
                     const Token *strtok = cback();
                     while (strtok->comment)
                         strtok = strtok->previous;
                     loc.push(location);
                     location.fileIndex = fileIndex(strtok->str().substr(1U, strtok->str().size() - 2U));
                     location.line = 1U;
-                } else if (lastline == "# line %num%") {
-                    const Token *numtok = cback();
-                    while (numtok->comment)
-                        numtok = numtok->previous;
-                    lineDirective(location.fileIndex, std::atol(numtok->str().c_str()), &location);
-                } else if (lastline == "# %num% %str%" || lastline == "# line %num% %str%") {
+                }
+                // #3 "file.c"
+                // #line 3 "file.c"
+                else if ((llNextToken->number &&
+                          llNextToken->next &&
+                          llNextToken->next->str()[0] == '\"') ||
+                         (llNextToken->str() == "line" &&
+                          llNextToken->next &&
+                          llNextToken->next->number &&
+                          llNextToken->next->next &&
+                          llNextToken->next->next->str()[0] == '\"'))
+                {
                     const Token *strtok = cback();
                     while (strtok->comment)
                         strtok = strtok->previous;
@@ -710,8 +725,19 @@ void simplecpp::TokenList::readfile(Stream &stream, const std::string &filename,
                     lineDirective(fileIndex(replaceAll(strtok->str().substr(1U, strtok->str().size() - 2U),"\\\\","\\")),
                                   std::atol(numtok->str().c_str()), &location);
                 }
+                // #line 3
+                else if (llNextToken->str() == "line" &&
+                         llNextToken->next &&
+                         llNextToken->next->number)
+                {
+                    const Token *numtok = cback();
+                    while (numtok->comment)
+                        numtok = numtok->previous;
+                    lineDirective(location.fileIndex, std::atol(numtok->str().c_str()), &location);
+                }
                 // #endfile
-                else if (lastline == "# endfile" && !loc.empty()) {
+                else if (llNextToken->str() == "endfile" && !loc.empty())
+                {
                     location = loc.top();
                     loc.pop();
                 }
@@ -1405,34 +1431,6 @@ std::string simplecpp::TokenList::readUntil(Stream &stream, const Location &loca
     return ret;
 }
 
-std::string simplecpp::TokenList::lastLine(int maxsize) const
-{
-    std::string ret;
-    int count = 0;
-    for (const Token *tok = cback(); ; tok = tok->previous) {
-        if (!sameline(tok, cback())) {
-            break;
-        }
-        if (tok->comment)
-            continue;
-        if (++count > maxsize)
-            return "";
-        if (!ret.empty())
-            ret += ' ';
-        // add tokens in reverse for performance reasons
-        if (tok->str()[0] == '\"')
-            ret += "%rts%"; // %str%
-        else if (tok->number)
-            ret += "%mun%"; // %num%
-        else {
-            ret += tok->str();
-            std::reverse(ret.end() - tok->str().length(), ret.end());
-        }
-    }
-    std::reverse(ret.begin(), ret.end());
-    return ret;
-}
-
 const simplecpp::Token* simplecpp::TokenList::lastLineTok(int maxsize) const
 {
     const Token* prevTok = nullptr;
@@ -1449,10 +1447,12 @@ const simplecpp::Token* simplecpp::TokenList::lastLineTok(int maxsize) const
     return prevTok;
 }
 
-bool simplecpp::TokenList::isLastLinePreprocessor(int maxsize) const
+const simplecpp::Token* simplecpp::TokenList::isLastLinePreprocessor(int maxsize) const
 {
     const Token * const prevTok = lastLineTok(maxsize);
-    return prevTok && prevTok->op == '#';
+    if (prevTok && prevTok->op == '#')
+        return prevTok;
+    return nullptr;
 }
 
 unsigned int simplecpp::TokenList::fileIndex(const std::string &filename)
@@ -1700,7 +1700,9 @@ namespace simplecpp {
             nameTokDef = nametoken;
             variadic = false;
             variadicOpt = false;
+            delete optExpandValue;
             optExpandValue = nullptr;
+            delete optNoExpandValue;
             optNoExpandValue = nullptr;
             if (!nameTokDef) {
                 valueToken = endToken = nullptr;
@@ -2374,8 +2376,8 @@ namespace simplecpp {
         bool variadicOpt;
 
         /** Expansion value for varadic macros with __VA_OPT__ expanded and discarded respectively */
-        const TokenList *optExpandValue;
-        const TokenList *optNoExpandValue;
+        const TokenList *optExpandValue = nullptr;
+        const TokenList *optNoExpandValue = nullptr;
 
         /** was the value of this macro actually defined in the code? */
         bool valueDefinedInCode_;
@@ -2979,9 +2981,11 @@ static std::string openHeaderDirect(std::ifstream &f, const std::string &path)
     if (nonExistingFilesCache.contains(path))
         return "";  // file is known not to exist, skip expensive file open call
 #endif
-    f.open(path.c_str());
-    if (f.is_open())
-        return path;
+    if (simplecpp::isFile(path)) {
+        f.open(path.c_str());
+        if (f.is_open())
+            return path;
+    }
 #ifdef SIMPLECPP_WINDOWS
     nonExistingFilesCache.add(path);
 #endif
@@ -3102,6 +3106,9 @@ bool simplecpp::FileDataCache::getFileId(const std::string &path, FileID &id)
     if (stat(path.c_str(), &statbuf) != 0)
         return false;
 
+    if ((statbuf.st_mode & S_IFMT) != S_IFREG)
+        return false;
+
     id.dev = statbuf.st_dev;
     id.ino = statbuf.st_ino;
 
@@ -3834,3 +3841,21 @@ std::string simplecpp::getCppStdString(const std::string &std)
 {
     return getCppStdString(getCppStd(std));
 }
+
+static mode_t file_type(const std::string &path)
+{
+    struct stat file_stat;
+    if (stat(path.c_str(), &file_stat) == -1)
+        return 0;
+    return file_stat.st_mode & S_IFMT;
+}
+
+bool simplecpp::isFile(const std::string &path)
+{
+    return file_type(path) == S_IFREG;
+}
+
+bool simplecpp::isDirectory(const std::string &path)
+{
+    return file_type(path) == S_IFDIR;
+}

simplecpp.h

@@ -310,9 +310,8 @@ namespace simplecpp {
         std::string readUntil(Stream &stream, const Location &location, char start, char end, OutputList *outputList);
         void lineDirective(unsigned int fileIndex, unsigned int line, Location *location);
 
-        std::string lastLine(int maxsize=1000) const;
         const Token* lastLineTok(int maxsize=1000) const;
-        bool isLastLinePreprocessor(int maxsize=1000) const;
+        const Token* isLastLinePreprocessor(int maxsize=1000) const;
 
         unsigned int fileIndex(const std::string &filename);
 
@@ -396,6 +395,20 @@ namespace simplecpp {
     SIMPLECPP_LIB std::string getCppStdString(const std::string &std);
     SIMPLECPP_LIB std::string getCppStdString(cppstd_t std);
 
+    /**
+     * @brief Checks if given path is a file
+     * @param path Path to be checked
+     * @return true if given path is a file
+     */
+    SIMPLECPP_LIB bool isFile(const std::string &path);
+
+    /**
+     * @brief Checks if a given path is a directory
+     * @param path Path to be checked
+     * @return true if given path is a directory
+     */
+    SIMPLECPP_LIB bool isDirectory(const std::string &path);
+
     struct SIMPLECPP_LIB FileData {
         /** The canonical filename associated with this data */
         std::string filename;