refactor: replace boost::filesystem with std::filesystem

Warning: Replacing fs::system_complete calls with fs::absolute calls
in this commit may cause minor changes in behaviour because fs::absolute
no longer strips trailing slashes; however these changes are believed to
be safe.

Co-authored-by: Russell Yanofsky <[email protected]>
Co-authored-by: Hennadii Stepanov <[email protected]>

Author: 3 people

src/addrdb.cpp

@@ -9,6 +9,7 @@
 #include <chainparams.h>
 #include <clientversion.h>
 #include <cstdint>
+#include <fs.h>
 #include <hash.h>
 #include <logging/timer.h>
 #include <netbase.h>

src/bench/bench.cpp

@@ -8,6 +8,7 @@
 #include <test/util/setup_common.h>
 
 #include <chrono>
+#include <fstream>
 #include <functional>
 #include <iostream>
 #include <map>
@@ -29,7 +30,7 @@ void GenerateTemplateResults(const std::vector<ankerl::nanobench::Result>& bench
         // nothing to write, bail out
         return;
     }
-    fsbridge::ofstream fout{fs::PathFromString(filename)};
+    std::ofstream fout{fs::PathFromString(filename)};
     if (fout.is_open()) {
         ankerl::nanobench::render(tpl, benchmarkResults, fout);
     } else {

src/fs.cpp

@@ -37,7 +37,7 @@ FILE *fopen(const fs::path& p, const char *mode)
 fs::path AbsPathJoin(const fs::path& base, const fs::path& path)
 {
     assert(base.is_absolute());
-    return fs::absolute(path, base);
+    return path.empty() ? base : fs::path(base / path);
 }
 
 #ifndef WIN32
@@ -153,118 +153,4 @@ std::string get_filesystem_error_message(const fs::filesystem_error& e)
 #endif
 }
 
-#ifdef WIN32
-#ifdef __GLIBCXX__
-
-// reference: https://github.com/gcc-mirror/gcc/blob/gcc-7_3_0-release/libstdc%2B%2B-v3/include/std/fstream#L270
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC diagnostic push
-#pragma GCC diagnostic ignored "-Wswitch"
-#endif
-static std::string openmodeToStr(std::ios_base::openmode mode)
-{
-    switch (mode & ~std::ios_base::ate) {
-    case std::ios_base::out:
-    case std::ios_base::out | std::ios_base::trunc:
-        return "w";
-    case std::ios_base::out | std::ios_base::app:
-    case std::ios_base::app:
-        return "a";
-    case std::ios_base::in:
-        return "r";
-    case std::ios_base::in | std::ios_base::out:
-        return "r+";
-    case std::ios_base::in | std::ios_base::out | std::ios_base::trunc:
-        return "w+";
-    case std::ios_base::in | std::ios_base::out | std::ios_base::app:
-    case std::ios_base::in | std::ios_base::app:
-        return "a+";
-    case std::ios_base::out | std::ios_base::binary:
-    case std::ios_base::out | std::ios_base::trunc | std::ios_base::binary:
-        return "wb";
-    case std::ios_base::out | std::ios_base::app | std::ios_base::binary:
-    case std::ios_base::app | std::ios_base::binary:
-        return "ab";
-    case std::ios_base::in | std::ios_base::binary:
-        return "rb";
-    case std::ios_base::in | std::ios_base::out | std::ios_base::binary:
-        return "r+b";
-    case std::ios_base::in | std::ios_base::out | std::ios_base::trunc | std::ios_base::binary:
-        return "w+b";
-    case std::ios_base::in | std::ios_base::out | std::ios_base::app | std::ios_base::binary:
-    case std::ios_base::in | std::ios_base::app | std::ios_base::binary:
-        return "a+b";
-    default:
-        return std::string();
-    }
-}
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC diagnostic pop
-#endif
-
-void ifstream::open(const fs::path& p, std::ios_base::openmode mode)
-{
-    close();
-    mode |= std::ios_base::in;
-    m_file = fsbridge::fopen(p, openmodeToStr(mode).c_str());
-    if (m_file == nullptr) {
-        return;
-    }
-    m_filebuf = __gnu_cxx::stdio_filebuf<char>(m_file, mode);
-    rdbuf(&m_filebuf);
-    if (mode & std::ios_base::ate) {
-        seekg(0, std::ios_base::end);
-    }
-}
-
-void ifstream::close()
-{
-    if (m_file != nullptr) {
-        m_filebuf.close();
-        fclose(m_file);
-    }
-    m_file = nullptr;
-}
-
-void ofstream::open(const fs::path& p, std::ios_base::openmode mode)
-{
-    close();
-    mode |= std::ios_base::out;
-    m_file = fsbridge::fopen(p, openmodeToStr(mode).c_str());
-    if (m_file == nullptr) {
-        return;
-    }
-    m_filebuf = __gnu_cxx::stdio_filebuf<char>(m_file, mode);
-    rdbuf(&m_filebuf);
-    if (mode & std::ios_base::ate) {
-        seekp(0, std::ios_base::end);
-    }
-}
-
-void ofstream::close()
-{
-    if (m_file != nullptr) {
-        m_filebuf.close();
-        fclose(m_file);
-    }
-    m_file = nullptr;
-}
-#else // __GLIBCXX__
-
-#if BOOST_VERSION >= 107700
-static_assert(sizeof(*BOOST_FILESYSTEM_C_STR(boost::filesystem::path())) == sizeof(wchar_t),
-#else
-static_assert(sizeof(*boost::filesystem::path().BOOST_FILESYSTEM_C_STR) == sizeof(wchar_t),
-#endif // BOOST_VERSION >= 107700
-    "Warning: This build is using boost::filesystem ofstream and ifstream "
-    "implementations which will fail to open paths containing multibyte "
-    "characters. You should delete this static_assert to ignore this warning, "
-    "or switch to a different C++ standard library like the Microsoft C++ "
-    "Standard Library (where boost uses non-standard extensions to construct "
-    "stream objects with wide filenames), or the GNU libstdc++ library (where "
-    "a more complicated workaround has been implemented above).");
-
-#endif // __GLIBCXX__
-#endif // WIN32
-
 } // fsbridge

src/fs.h

@@ -5,46 +5,42 @@
 #ifndef BITCOIN_FS_H
 #define BITCOIN_FS_H
 
-#include <stdio.h>
-#include <string>
-#if defined WIN32 && defined __GLIBCXX__
-#include <ext/stdio_filebuf.h>
-#endif
-
-#include <boost/filesystem.hpp>
-#include <boost/filesystem/fstream.hpp>
 #include <tinyformat.h>
 
+#include <cstdio>
+#include <filesystem>
+#include <iomanip>
+#include <ios>
+#include <ostream>
+#include <string>
+#include <utility>
+
 /** Filesystem operations and types */
 namespace fs {
 
-using namespace boost::filesystem;
+using namespace std::filesystem;
 
 /**
- * Path class wrapper to prepare application code for transition from
- * boost::filesystem library to std::filesystem implementation. The main
- * purpose of the class is to define fs::path::u8string() and fs::u8path()
- * functions not present in boost. It also blocks calls to the
- * fs::path(std::string) implicit constructor and the fs::path::string()
- * method, which worked well in the boost::filesystem implementation, but have
- * unsafe and unpredictable behavior on Windows in the std::filesystem
- * implementation (see implementation note in \ref PathToString for details).
+ * Path class wrapper to block calls to the fs::path(std::string) implicit
+ * constructor and the fs::path::string() method, which have unsafe and
+ * unpredictable behavior on Windows (see implementation note in
+ * \ref PathToString for details)
  */
-class path : public boost::filesystem::path
+class path : public std::filesystem::path
 {
 public:
-    using boost::filesystem::path::path;
+    using std::filesystem::path::path;
 
     // Allow path objects arguments for compatibility.
-    path(boost::filesystem::path path) : boost::filesystem::path::path(std::move(path)) {}
-    path& operator=(boost::filesystem::path path) { boost::filesystem::path::operator=(std::move(path)); return *this; }
-    path& operator/=(boost::filesystem::path path) { boost::filesystem::path::operator/=(std::move(path)); return *this; }
+    path(std::filesystem::path path) : std::filesystem::path::path(std::move(path)) {}
+    path& operator=(std::filesystem::path path) { std::filesystem::path::operator=(std::move(path)); return *this; }
+    path& operator/=(std::filesystem::path path) { std::filesystem::path::operator/=(std::move(path)); return *this; }
 
     // Allow literal string arguments, which are safe as long as the literals are ASCII.
-    path(const char* c) : boost::filesystem::path(c) {}
-    path& operator=(const char* c) { boost::filesystem::path::operator=(c); return *this; }
-    path& operator/=(const char* c) { boost::filesystem::path::operator/=(c); return *this; }
-    path& append(const char* c) { boost::filesystem::path::append(c); return *this; }
+    path(const char* c) : std::filesystem::path(c) {}
+    path& operator=(const char* c) { std::filesystem::path::operator=(c); return *this; }
+    path& operator/=(const char* c) { std::filesystem::path::operator/=(c); return *this; }
+    path& append(const char* c) { std::filesystem::path::append(c); return *this; }
 
     // Disallow std::string arguments to avoid locale-dependent decoding on windows.
     path(std::string) = delete;
@@ -55,52 +51,48 @@ class path : public boost::filesystem::path
     // Disallow std::string conversion method to avoid locale-dependent encoding on windows.
     std::string string() const = delete;
 
-    // Define UTF-8 string conversion method not present in boost::filesystem but present in std::filesystem.
-    std::string u8string() const { return boost::filesystem::path::string(); }
+    // Required for path overloads in <fstream>.
+    // See https://gcc.gnu.org/git/?p=gcc.git;a=commit;h=96e0367ead5d8dcac3bec2865582e76e2fbab190
+    path& make_preferred() { std::filesystem::path::make_preferred(); return *this; }
+    path filename() const { return std::filesystem::path::filename(); }
 };
 
-// Define UTF-8 string conversion function not present in boost::filesystem but present in std::filesystem.
-static inline path u8path(const std::string& string)
-{
-    return boost::filesystem::path(string);
-}
-
-// Disallow implicit std::string conversion for system_complete to avoid
+// Disallow implicit std::string conversion for absolute to avoid
 // locale-dependent encoding on windows.
-static inline path system_complete(const path& p)
+static inline path absolute(const path& p)
 {
-    return boost::filesystem::system_complete(p);
+    return std::filesystem::absolute(p);
 }
 
 // Disallow implicit std::string conversion for exists to avoid
 // locale-dependent encoding on windows.
 static inline bool exists(const path& p)
 {
-    return boost::filesystem::exists(p);
+    return std::filesystem::exists(p);
 }
 
 // Allow explicit quoted stream I/O.
 static inline auto quoted(const std::string& s)
 {
-    return boost::io::quoted(s, '&');
+    return std::quoted(s, '"', '&');
 }
 
 // Allow safe path append operations.
 static inline path operator+(path p1, path p2)
 {
-    p1 += static_cast<boost::filesystem::path&&>(p2);
+    p1 += std::move(p2);
     return p1;
 }
 
 // Disallow implicit std::string conversion for copy_file
 // to avoid locale-dependent encoding on Windows.
-static inline void copy_file(const path& from, const path& to, copy_option options)
+static inline bool copy_file(const path& from, const path& to, copy_options options)
 {
-    boost::filesystem::copy_file(from, to, options);
+    return std::filesystem::copy_file(from, to, options);
 }
 
 /**
- * Convert path object to byte string. On POSIX, paths natively are byte
+ * Convert path object to a byte string. On POSIX, paths natively are byte
  * strings, so this is trivial. On Windows, paths natively are Unicode, so an
  * encoding step is necessary. The inverse of \ref PathToString is \ref
  * PathFromString. The strings returned and parsed by these functions can be
@@ -112,7 +104,7 @@ static inline void copy_file(const path& from, const path& to, copy_option optio
  * appropriate to use in applications requiring UTF-8, where
  * fs::path::u8string() and fs::u8path() methods should be used instead. Other
  * applications could require still different encodings. For example, JSON, XML,
- * or URI applications might prefer to use higher level escapes (\uXXXX or
+ * or URI applications might prefer to use higher-level escapes (\uXXXX or
  * &XXXX; or %XX) instead of multibyte encoding. Rust, Python, Java applications
  * may require encoding paths with their respective UTF-8 derivatives WTF-8,
  * PEP-383, and CESU-8 (see https://en.wikipedia.org/wiki/UTF-8#Derivatives).
@@ -133,7 +125,7 @@ static inline std::string PathToString(const path& path)
     return path.u8string();
 #else
     static_assert(std::is_same<path::string_type, std::string>::value, "PathToString not implemented on this platform");
-    return path.boost::filesystem::path::string();
+    return path.std::filesystem::path::string();
 #endif
 }
 
@@ -145,7 +137,7 @@ static inline path PathFromString(const std::string& string)
 #ifdef WIN32
     return u8path(string);
 #else
-    return boost::filesystem::path(string);
+    return std::filesystem::path(string);
 #endif
 }
 } // namespace fs
@@ -186,60 +178,12 @@ namespace fsbridge {
     };
 
     std::string get_filesystem_error_message(const fs::filesystem_error& e);
-
-    // GNU libstdc++ specific workaround for opening UTF-8 paths on Windows.
-    //
-    // On Windows, it is only possible to reliably access multibyte file paths through
-    // `wchar_t` APIs, not `char` APIs. But because the C++ standard doesn't
-    // require ifstream/ofstream `wchar_t` constructors, and the GNU library doesn't
-    // provide them (in contrast to the Microsoft C++ library, see
-    // https://stackoverflow.com/questions/821873/how-to-open-an-stdfstream-ofstream-or-ifstream-with-a-unicode-filename/822032#822032),
-    // Boost is forced to fall back to `char` constructors which may not work properly.
-    //
-    // Work around this issue by creating stream objects with `_wfopen` in
-    // combination with `__gnu_cxx::stdio_filebuf`. This workaround can be removed
-    // with an upgrade to C++17, where streams can be constructed directly from
-    // `std::filesystem::path` objects.
-
-#if defined WIN32 && defined __GLIBCXX__
-    class ifstream : public std::istream
-    {
-    public:
-        ifstream() = default;
-        explicit ifstream(const fs::path& p, std::ios_base::openmode mode = std::ios_base::in) { open(p, mode); }
-        ~ifstream() { close(); }
-        void open(const fs::path& p, std::ios_base::openmode mode = std::ios_base::in);
-        bool is_open() { return m_filebuf.is_open(); }
-        void close();
-
-    private:
-        __gnu_cxx::stdio_filebuf<char> m_filebuf;
-        FILE* m_file = nullptr;
-    };
-    class ofstream : public std::ostream
-    {
-    public:
-        ofstream() = default;
-        explicit ofstream(const fs::path& p, std::ios_base::openmode mode = std::ios_base::out) { open(p, mode); }
-        ~ofstream() { close(); }
-        void open(const fs::path& p, std::ios_base::openmode mode = std::ios_base::out);
-        bool is_open() { return m_filebuf.is_open(); }
-        void close();
-
-    private:
-        __gnu_cxx::stdio_filebuf<char> m_filebuf;
-        FILE* m_file = nullptr;
-    };
-#else  // !(WIN32 && __GLIBCXX__)
-    typedef fs::ifstream ifstream;
-    typedef fs::ofstream ofstream;
-#endif // WIN32 && __GLIBCXX__
 };
 
 // Disallow path operator<< formatting in tinyformat to avoid locale-dependent
 // encoding on windows.
 namespace tinyformat {
-template<> inline void formatValue(std::ostream&, const char*, const char*, int, const boost::filesystem::path&) = delete;
+template<> inline void formatValue(std::ostream&, const char*, const char*, int, const std::filesystem::path&) = delete;
 template<> inline void formatValue(std::ostream&, const char*, const char*, int, const fs::path&) = delete;
 } // namespace tinyformat