diff --git a/thirdparty/CLI11.hpp b/thirdparty/CLI11.hpp index 3913fa9c..41027f0e 100644 --- a/thirdparty/CLI11.hpp +++ b/thirdparty/CLI11.hpp @@ -1,11 +1,11 @@ -// CLI11: Version 2.3.2 +// CLI11: Version 2.4.1 // Originally designed by Henry Schreiner // https://github.com/CLIUtils/CLI11 // // This is a standalone header file generated by MakeSingleHeader.py in CLI11/scripts -// from: v2.3.2 +// from: v2.4.1 // -// CLI11 2.3.2 Copyright (c) 2017-2022 University of Cincinnati, developed by Henry +// CLI11 2.4.1 Copyright (c) 2017-2024 University of Cincinnati, developed by Henry // Schreiner under NSF AWARD 1414736. All rights reserved. // // Redistribution and use in source and binary forms of CLI11, with or without @@ -34,34 +34,40 @@ #pragma once // Standard combined includes: -#include -#include -#include -#include -#include -#include -#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include #include +#include #include #include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include #include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include #define CLI11_VERSION_MAJOR 2 -#define CLI11_VERSION_MINOR 3 -#define CLI11_VERSION_PATCH 2 -#define CLI11_VERSION "2.3.2" +#define CLI11_VERSION_MINOR 4 +#define CLI11_VERSION_PATCH 1 +#define CLI11_VERSION "2.4.1" @@ -124,17 +130,7 @@ #endif #endif -/** Inline macro **/ -#ifdef CLI11_COMPILE -#define CLI11_INLINE -#else -#define CLI11_INLINE inline -#endif - - - -// C standard library -// Only needed for existence checking +/** availability */ #if defined CLI11_CPP17 && defined __has_include && !defined CLI11_HAS_FILESYSTEM #if __has_include() // Filesystem cannot be used if targeting macOS < 10.15 @@ -161,6 +157,44 @@ #endif #endif +/** availability */ +#if defined(__GNUC__) && !defined(__llvm__) && !defined(__INTEL_COMPILER) && __GNUC__ < 5 +#define CLI11_HAS_CODECVT 0 +#else +#define CLI11_HAS_CODECVT 1 +#include +#endif + +/** disable deprecations */ +#if defined(__GNUC__) // GCC or clang +#define CLI11_DIAGNOSTIC_PUSH _Pragma("GCC diagnostic push") +#define CLI11_DIAGNOSTIC_POP _Pragma("GCC diagnostic pop") + +#define CLI11_DIAGNOSTIC_IGNORE_DEPRECATED _Pragma("GCC diagnostic ignored \"-Wdeprecated-declarations\"") + +#elif defined(_MSC_VER) +#define CLI11_DIAGNOSTIC_PUSH __pragma(warning(push)) +#define CLI11_DIAGNOSTIC_POP __pragma(warning(pop)) + +#define CLI11_DIAGNOSTIC_IGNORE_DEPRECATED __pragma(warning(disable : 4996)) + +#else +#define CLI11_DIAGNOSTIC_PUSH +#define CLI11_DIAGNOSTIC_POP + +#define CLI11_DIAGNOSTIC_IGNORE_DEPRECATED + +#endif + +/** Inline macro **/ +#ifdef CLI11_COMPILE +#define CLI11_INLINE +#else +#define CLI11_INLINE inline +#endif + + + #if defined CLI11_HAS_FILESYSTEM && CLI11_HAS_FILESYSTEM > 0 #include // NOLINT(build/include) #else @@ -170,9 +204,244 @@ + +#ifdef CLI11_CPP17 +#include +#endif // CLI11_CPP17 + +#if defined CLI11_HAS_FILESYSTEM && CLI11_HAS_FILESYSTEM > 0 +#include +#include // NOLINT(build/include) +#endif // CLI11_HAS_FILESYSTEM + + + +#if defined(_WIN32) +#if !(defined(_AMD64_) || defined(_X86_) || defined(_ARM_)) +#if defined(__amd64__) || defined(__amd64) || defined(__x86_64__) || defined(__x86_64) || defined(_M_X64) || \ + defined(_M_AMD64) +#define _AMD64_ +#elif defined(i386) || defined(__i386) || defined(__i386__) || defined(__i386__) || defined(_M_IX86) +#define _X86_ +#elif defined(__arm__) || defined(_M_ARM) || defined(_M_ARMT) +#define _ARM_ +#elif defined(__aarch64__) || defined(_M_ARM64) +#define _ARM64_ +#elif defined(_M_ARM64EC) +#define _ARM64EC_ +#endif +#endif + +// first +#ifndef NOMINMAX +// if NOMINMAX is already defined we don't want to mess with that either way +#define NOMINMAX +#include +#undef NOMINMAX +#else +#include +#endif + +// second +#include +// third +#include +#include +#endif + + namespace CLI { +/// Convert a wide string to a narrow string. +CLI11_INLINE std::string narrow(const std::wstring &str); +CLI11_INLINE std::string narrow(const wchar_t *str); +CLI11_INLINE std::string narrow(const wchar_t *str, std::size_t size); + +/// Convert a narrow string to a wide string. +CLI11_INLINE std::wstring widen(const std::string &str); +CLI11_INLINE std::wstring widen(const char *str); +CLI11_INLINE std::wstring widen(const char *str, std::size_t size); + +#ifdef CLI11_CPP17 +CLI11_INLINE std::string narrow(std::wstring_view str); +CLI11_INLINE std::wstring widen(std::string_view str); +#endif // CLI11_CPP17 + +#if defined CLI11_HAS_FILESYSTEM && CLI11_HAS_FILESYSTEM > 0 +/// Convert a char-string to a native path correctly. +CLI11_INLINE std::filesystem::path to_path(std::string_view str); +#endif // CLI11_HAS_FILESYSTEM + + + + +namespace detail { + +#if !CLI11_HAS_CODECVT +/// Attempt to set one of the acceptable unicode locales for conversion +CLI11_INLINE void set_unicode_locale() { + static const std::array unicode_locales{{"C.UTF-8", "en_US.UTF-8", ".UTF-8"}}; + + for(const auto &locale_name : unicode_locales) { + if(std::setlocale(LC_ALL, locale_name) != nullptr) { + return; + } + } + throw std::runtime_error("CLI::narrow: could not set locale to C.UTF-8"); +} + +template struct scope_guard_t { + F closure; + + explicit scope_guard_t(F closure_) : closure(closure_) {} + ~scope_guard_t() { closure(); } +}; + +template CLI11_NODISCARD CLI11_INLINE scope_guard_t scope_guard(F &&closure) { + return scope_guard_t{std::forward(closure)}; +} + +#endif // !CLI11_HAS_CODECVT + +CLI11_DIAGNOSTIC_PUSH +CLI11_DIAGNOSTIC_IGNORE_DEPRECATED + +CLI11_INLINE std::string narrow_impl(const wchar_t *str, std::size_t str_size) { +#if CLI11_HAS_CODECVT +#ifdef _WIN32 + return std::wstring_convert>().to_bytes(str, str + str_size); + +#else + return std::wstring_convert>().to_bytes(str, str + str_size); + +#endif // _WIN32 +#else // CLI11_HAS_CODECVT + (void)str_size; + std::mbstate_t state = std::mbstate_t(); + const wchar_t *it = str; + + std::string old_locale = std::setlocale(LC_ALL, nullptr); + auto sg = scope_guard([&] { std::setlocale(LC_ALL, old_locale.c_str()); }); + set_unicode_locale(); + + std::size_t new_size = std::wcsrtombs(nullptr, &it, 0, &state); + if(new_size == static_cast(-1)) { + throw std::runtime_error("CLI::narrow: conversion error in std::wcsrtombs at offset " + + std::to_string(it - str)); + } + std::string result(new_size, '\0'); + std::wcsrtombs(const_cast(result.data()), &str, new_size, &state); + + return result; + +#endif // CLI11_HAS_CODECVT +} + +CLI11_INLINE std::wstring widen_impl(const char *str, std::size_t str_size) { +#if CLI11_HAS_CODECVT +#ifdef _WIN32 + return std::wstring_convert>().from_bytes(str, str + str_size); + +#else + return std::wstring_convert>().from_bytes(str, str + str_size); + +#endif // _WIN32 +#else // CLI11_HAS_CODECVT + (void)str_size; + std::mbstate_t state = std::mbstate_t(); + const char *it = str; + + std::string old_locale = std::setlocale(LC_ALL, nullptr); + auto sg = scope_guard([&] { std::setlocale(LC_ALL, old_locale.c_str()); }); + set_unicode_locale(); + + std::size_t new_size = std::mbsrtowcs(nullptr, &it, 0, &state); + if(new_size == static_cast(-1)) { + throw std::runtime_error("CLI::widen: conversion error in std::mbsrtowcs at offset " + + std::to_string(it - str)); + } + std::wstring result(new_size, L'\0'); + std::mbsrtowcs(const_cast(result.data()), &str, new_size, &state); + + return result; + +#endif // CLI11_HAS_CODECVT +} + +CLI11_DIAGNOSTIC_POP + +} // namespace detail + +CLI11_INLINE std::string narrow(const wchar_t *str, std::size_t str_size) { return detail::narrow_impl(str, str_size); } +CLI11_INLINE std::string narrow(const std::wstring &str) { return detail::narrow_impl(str.data(), str.size()); } +// Flawfinder: ignore +CLI11_INLINE std::string narrow(const wchar_t *str) { return detail::narrow_impl(str, std::wcslen(str)); } + +CLI11_INLINE std::wstring widen(const char *str, std::size_t str_size) { return detail::widen_impl(str, str_size); } +CLI11_INLINE std::wstring widen(const std::string &str) { return detail::widen_impl(str.data(), str.size()); } +// Flawfinder: ignore +CLI11_INLINE std::wstring widen(const char *str) { return detail::widen_impl(str, std::strlen(str)); } + +#ifdef CLI11_CPP17 +CLI11_INLINE std::string narrow(std::wstring_view str) { return detail::narrow_impl(str.data(), str.size()); } +CLI11_INLINE std::wstring widen(std::string_view str) { return detail::widen_impl(str.data(), str.size()); } +#endif // CLI11_CPP17 + +#if defined CLI11_HAS_FILESYSTEM && CLI11_HAS_FILESYSTEM > 0 +CLI11_INLINE std::filesystem::path to_path(std::string_view str) { + return std::filesystem::path{ +#ifdef _WIN32 + widen(str) +#else + str +#endif // _WIN32 + }; +} +#endif // CLI11_HAS_FILESYSTEM + + + + +namespace detail { +#ifdef _WIN32 +/// Decode and return UTF-8 argv from GetCommandLineW. +CLI11_INLINE std::vector compute_win32_argv(); +#endif +} // namespace detail + + + +namespace detail { + +#ifdef _WIN32 +CLI11_INLINE std::vector compute_win32_argv() { + std::vector result; + int argc = 0; + + auto deleter = [](wchar_t **ptr) { LocalFree(ptr); }; + // NOLINTBEGIN(*-avoid-c-arrays) + auto wargv = std::unique_ptr(CommandLineToArgvW(GetCommandLineW(), &argc), deleter); + // NOLINTEND(*-avoid-c-arrays) + + if(wargv == nullptr) { + throw std::runtime_error("CommandLineToArgvW failed with code " + std::to_string(GetLastError())); + } + + result.reserve(static_cast(argc)); + for(size_t i = 0; i < static_cast(argc); ++i) { + result.push_back(narrow(wargv[i])); + } + + return result; +} +#endif + +} // namespace detail + + + + /// Include the items in this namespace to get free conversion of enums to/from streams. /// (This is available inside CLI as well, so CLI11 will use this without a using statement). namespace enums { @@ -270,6 +539,9 @@ inline std::string trim_copy(const std::string &str) { /// remove quotes at the front and back of a string either '"' or '\'' CLI11_INLINE std::string &remove_quotes(std::string &str); +/// remove quotes from all elements of a string vector and process escaped components +CLI11_INLINE void remove_quotes(std::vector &args); + /// Add a leader to the beginning of all new lines (nothing is added /// at the start of the first line). `"; "` would be for ini files /// @@ -290,14 +562,16 @@ CLI11_INLINE std::ostream &format_aliases(std::ostream &out, const std::vector bool valid_first_char(T c) { return ((c != '-') && (c != '!') && (c != ' ') && c != '\n'); } +template bool valid_first_char(T c) { + return ((c != '-') && (static_cast(c) > 33)); // space and '!' not allowed +} /// Verify following characters of an option template bool valid_later_char(T c) { // = and : are value separators, { has special meaning for option defaults, - // and \n would just be annoying to deal with in many places allowing space here has too much potential for - // inadvertent entry errors and bugs - return ((c != '=') && (c != ':') && (c != '{') && (c != ' ') && c != '\n'); + // and control codes other than tab would just be annoying to deal with in many places allowing space here has too + // much potential for inadvertent entry errors and bugs + return ((c != '=') && (c != ':') && (c != '{') && ((static_cast(c) > 32) || c == '\t')); } /// Verify an option/subcommand name @@ -360,18 +634,46 @@ template inline std::string find_and_modify(std::string str, return str; } +/// close a sequence of characters indicated by a closure character. Brackets allows sub sequences +/// recognized bracket sequences include "'`[(<{ other closure characters are assumed to be literal strings +CLI11_INLINE std::size_t close_sequence(const std::string &str, std::size_t start, char closure_char); + /// Split a string '"one two" "three"' into 'one two', 'three' -/// Quote characters can be ` ' or " +/// Quote characters can be ` ' or " or bracket characters [{(< with matching to the matching bracket CLI11_INLINE std::vector split_up(std::string str, char delimiter = '\0'); +/// get the value of an environmental variable or empty string if empty +CLI11_INLINE std::string get_environment_value(const std::string &env_name); + /// This function detects an equal or colon followed by an escaped quote after an argument /// then modifies the string to replace the equality with a space. This is needed /// to allow the split up function to work properly and is intended to be used with the find_and_modify function /// the return value is the offset+1 which is required by the find_and_modify function. CLI11_INLINE std::size_t escape_detect(std::string &str, std::size_t offset); -/// Add quotes if the string contains spaces -CLI11_INLINE std::string &add_quotes_if_needed(std::string &str); +/// @brief detect if a string has escapable characters +/// @param str the string to do the detection on +/// @return true if the string has escapable characters +CLI11_INLINE bool has_escapable_character(const std::string &str); + +/// @brief escape all escapable characters +/// @param str the string to escape +/// @return a string with the escapble characters escaped with '\' +CLI11_INLINE std::string add_escaped_characters(const std::string &str); + +/// @brief replace the escaped characters with their equivalent +CLI11_INLINE std::string remove_escaped_characters(const std::string &str); + +/// generate a string with all non printable characters escaped to hex codes +CLI11_INLINE std::string binary_escape_string(const std::string &string_to_escape); + +CLI11_INLINE bool is_binary_escaped_string(const std::string &escaped_string); + +/// extract an escaped binary_string +CLI11_INLINE std::string extract_binary_string(const std::string &escaped_string); + +/// process a quoted string, remove the quotes and if appropriate handle escaped characters +CLI11_INLINE bool process_quoted_string(std::string &str, char string_char = '\"', char literal_char = '\''); } // namespace detail @@ -421,7 +723,17 @@ CLI11_INLINE std::string &rtrim(std::string &str, const std::string &filter) { } CLI11_INLINE std::string &remove_quotes(std::string &str) { - if(str.length() > 1 && (str.front() == '"' || str.front() == '\'')) { + if(str.length() > 1 && (str.front() == '"' || str.front() == '\'' || str.front() == '`')) { + if(str.front() == str.back()) { + str.pop_back(); + str.erase(str.begin(), str.begin() + 1); + } + } + return str; +} + +CLI11_INLINE std::string &remove_outer(std::string &str, char key) { + if(str.length() > 1 && (str.front() == key)) { if(str.front() == str.back()) { str.pop_back(); str.erase(str.begin(), str.begin() + 1); @@ -541,37 +853,220 @@ find_member(std::string name, const std::vector names, bool ignore_ return (it != std::end(names)) ? (it - std::begin(names)) : (-1); } +static const std::string escapedChars("\b\t\n\f\r\"\\"); +static const std::string escapedCharsCode("btnfr\"\\"); +static const std::string bracketChars{"\"'`[(<{"}; +static const std::string matchBracketChars("\"'`])>}"); + +CLI11_INLINE bool has_escapable_character(const std::string &str) { + return (str.find_first_of(escapedChars) != std::string::npos); +} + +CLI11_INLINE std::string add_escaped_characters(const std::string &str) { + std::string out; + out.reserve(str.size() + 4); + for(char s : str) { + auto sloc = escapedChars.find_first_of(s); + if(sloc != std::string::npos) { + out.push_back('\\'); + out.push_back(escapedCharsCode[sloc]); + } else { + out.push_back(s); + } + } + return out; +} + +CLI11_INLINE std::uint32_t hexConvert(char hc) { + int hcode{0}; + if(hc >= '0' && hc <= '9') { + hcode = (hc - '0'); + } else if(hc >= 'A' && hc <= 'F') { + hcode = (hc - 'A' + 10); + } else if(hc >= 'a' && hc <= 'f') { + hcode = (hc - 'a' + 10); + } else { + hcode = -1; + } + return static_cast(hcode); +} + +CLI11_INLINE char make_char(std::uint32_t code) { return static_cast(static_cast(code)); } + +CLI11_INLINE void append_codepoint(std::string &str, std::uint32_t code) { + if(code < 0x80) { // ascii code equivalent + str.push_back(static_cast(code)); + } else if(code < 0x800) { // \u0080 to \u07FF + // 110yyyyx 10xxxxxx; 0x3f == 0b0011'1111 + str.push_back(make_char(0xC0 | code >> 6)); + str.push_back(make_char(0x80 | (code & 0x3F))); + } else if(code < 0x10000) { // U+0800...U+FFFF + if(0xD800 <= code && code <= 0xDFFF) { + throw std::invalid_argument("[0xD800, 0xDFFF] are not valid UTF-8."); + } + // 1110yyyy 10yxxxxx 10xxxxxx + str.push_back(make_char(0xE0 | code >> 12)); + str.push_back(make_char(0x80 | (code >> 6 & 0x3F))); + str.push_back(make_char(0x80 | (code & 0x3F))); + } else if(code < 0x110000) { // U+010000 ... U+10FFFF + // 11110yyy 10yyxxxx 10xxxxxx 10xxxxxx + str.push_back(make_char(0xF0 | code >> 18)); + str.push_back(make_char(0x80 | (code >> 12 & 0x3F))); + str.push_back(make_char(0x80 | (code >> 6 & 0x3F))); + str.push_back(make_char(0x80 | (code & 0x3F))); + } +} + +CLI11_INLINE std::string remove_escaped_characters(const std::string &str) { + + std::string out; + out.reserve(str.size()); + for(auto loc = str.begin(); loc < str.end(); ++loc) { + if(*loc == '\\') { + if(str.end() - loc < 2) { + throw std::invalid_argument("invalid escape sequence " + str); + } + auto ecloc = escapedCharsCode.find_first_of(*(loc + 1)); + if(ecloc != std::string::npos) { + out.push_back(escapedChars[ecloc]); + ++loc; + } else if(*(loc + 1) == 'u') { + // must have 4 hex characters + if(str.end() - loc < 6) { + throw std::invalid_argument("unicode sequence must have 4 hex codes " + str); + } + std::uint32_t code{0}; + std::uint32_t mplier{16 * 16 * 16}; + for(int ii = 2; ii < 6; ++ii) { + std::uint32_t res = hexConvert(*(loc + ii)); + if(res > 0x0F) { + throw std::invalid_argument("unicode sequence must have 4 hex codes " + str); + } + code += res * mplier; + mplier = mplier / 16; + } + append_codepoint(out, code); + loc += 5; + } else if(*(loc + 1) == 'U') { + // must have 8 hex characters + if(str.end() - loc < 10) { + throw std::invalid_argument("unicode sequence must have 8 hex codes " + str); + } + std::uint32_t code{0}; + std::uint32_t mplier{16 * 16 * 16 * 16 * 16 * 16 * 16}; + for(int ii = 2; ii < 10; ++ii) { + std::uint32_t res = hexConvert(*(loc + ii)); + if(res > 0x0F) { + throw std::invalid_argument("unicode sequence must have 8 hex codes " + str); + } + code += res * mplier; + mplier = mplier / 16; + } + append_codepoint(out, code); + loc += 9; + } else if(*(loc + 1) == '0') { + out.push_back('\0'); + ++loc; + } else { + throw std::invalid_argument(std::string("unrecognized escape sequence \\") + *(loc + 1) + " in " + str); + } + } else { + out.push_back(*loc); + } + } + return out; +} + +CLI11_INLINE std::size_t close_string_quote(const std::string &str, std::size_t start, char closure_char) { + std::size_t loc{0}; + for(loc = start + 1; loc < str.size(); ++loc) { + if(str[loc] == closure_char) { + break; + } + if(str[loc] == '\\') { + // skip the next character for escaped sequences + ++loc; + } + } + return loc; +} + +CLI11_INLINE std::size_t close_literal_quote(const std::string &str, std::size_t start, char closure_char) { + auto loc = str.find_first_of(closure_char, start + 1); + return (loc != std::string::npos ? loc : str.size()); +} + +CLI11_INLINE std::size_t close_sequence(const std::string &str, std::size_t start, char closure_char) { + + auto bracket_loc = matchBracketChars.find(closure_char); + switch(bracket_loc) { + case 0: + return close_string_quote(str, start, closure_char); + case 1: + case 2: + case std::string::npos: + return close_literal_quote(str, start, closure_char); + default: + break; + } + + std::string closures(1, closure_char); + auto loc = start + 1; + + while(loc < str.size()) { + if(str[loc] == closures.back()) { + closures.pop_back(); + if(closures.empty()) { + return loc; + } + } + bracket_loc = bracketChars.find(str[loc]); + if(bracket_loc != std::string::npos) { + switch(bracket_loc) { + case 0: + loc = close_string_quote(str, loc, str[loc]); + break; + case 1: + case 2: + loc = close_literal_quote(str, loc, str[loc]); + break; + default: + closures.push_back(matchBracketChars[bracket_loc]); + break; + } + } + ++loc; + } + if(loc > str.size()) { + loc = str.size(); + } + return loc; +} + CLI11_INLINE std::vector split_up(std::string str, char delimiter) { - const std::string delims("\'\"`"); auto find_ws = [delimiter](char ch) { return (delimiter == '\0') ? std::isspace(ch, std::locale()) : (ch == delimiter); }; trim(str); std::vector output; - bool embeddedQuote = false; - char keyChar = ' '; while(!str.empty()) { - if(delims.find_first_of(str[0]) != std::string::npos) { - keyChar = str[0]; - auto end = str.find_first_of(keyChar, 1); - while((end != std::string::npos) && (str[end - 1] == '\\')) { // deal with escaped quotes - end = str.find_first_of(keyChar, end + 1); - embeddedQuote = true; - } - if(end != std::string::npos) { - output.push_back(str.substr(1, end - 1)); + if(bracketChars.find_first_of(str[0]) != std::string::npos) { + auto bracketLoc = bracketChars.find_first_of(str[0]); + auto end = close_sequence(str, 0, matchBracketChars[bracketLoc]); + if(end >= str.size()) { + output.push_back(std::move(str)); + str.clear(); + } else { + output.push_back(str.substr(0, end + 1)); if(end + 2 < str.size()) { str = str.substr(end + 2); } else { str.clear(); } - - } else { - output.push_back(str.substr(1)); - str = ""; } + } else { auto it = std::find_if(std::begin(str), std::end(str), find_ws); if(it != std::end(str)) { @@ -580,14 +1075,9 @@ CLI11_INLINE std::vector split_up(std::string str, char delimiter) str = std::string(it + 1, str.end()); } else { output.push_back(str); - str = ""; + str.clear(); } } - // transform any embedded quotes into the regular character - if(embeddedQuote) { - output.back() = find_and_replace(output.back(), std::string("\\") + keyChar, std::string(1, keyChar)); - embeddedQuote = false; - } trim(str); } return output; @@ -605,15 +1095,140 @@ CLI11_INLINE std::size_t escape_detect(std::string &str, std::size_t offset) { return offset + 1; } -CLI11_INLINE std::string &add_quotes_if_needed(std::string &str) { - if((str.front() != '"' && str.front() != '\'') || str.front() != str.back()) { - char quote = str.find('"') < str.find('\'') ? '\'' : '"'; - if(str.find(' ') != std::string::npos) { - str.insert(0, 1, quote); - str.append(1, quote); +CLI11_INLINE std::string binary_escape_string(const std::string &string_to_escape) { + // s is our escaped output string + std::string escaped_string{}; + // loop through all characters + for(char c : string_to_escape) { + // check if a given character is printable + // the cast is necessary to avoid undefined behaviour + if(isprint(static_cast(c)) == 0) { + std::stringstream stream; + // if the character is not printable + // we'll convert it to a hex string using a stringstream + // note that since char is signed we have to cast it to unsigned first + stream << std::hex << static_cast(static_cast(c)); + std::string code = stream.str(); + escaped_string += std::string("\\x") + (code.size() < 2 ? "0" : "") + code; + + } else { + escaped_string.push_back(c); } } - return str; + if(escaped_string != string_to_escape) { + auto sqLoc = escaped_string.find('\''); + while(sqLoc != std::string::npos) { + escaped_string.replace(sqLoc, sqLoc + 1, "\\x27"); + sqLoc = escaped_string.find('\''); + } + escaped_string.insert(0, "'B\"("); + escaped_string.push_back(')'); + escaped_string.push_back('"'); + escaped_string.push_back('\''); + } + return escaped_string; +} + +CLI11_INLINE bool is_binary_escaped_string(const std::string &escaped_string) { + size_t ssize = escaped_string.size(); + if(escaped_string.compare(0, 3, "B\"(") == 0 && escaped_string.compare(ssize - 2, 2, ")\"") == 0) { + return true; + } + return (escaped_string.compare(0, 4, "'B\"(") == 0 && escaped_string.compare(ssize - 3, 3, ")\"'") == 0); +} + +CLI11_INLINE std::string extract_binary_string(const std::string &escaped_string) { + std::size_t start{0}; + std::size_t tail{0}; + size_t ssize = escaped_string.size(); + if(escaped_string.compare(0, 3, "B\"(") == 0 && escaped_string.compare(ssize - 2, 2, ")\"") == 0) { + start = 3; + tail = 2; + } else if(escaped_string.compare(0, 4, "'B\"(") == 0 && escaped_string.compare(ssize - 3, 3, ")\"'") == 0) { + start = 4; + tail = 3; + } + + if(start == 0) { + return escaped_string; + } + std::string outstring; + + outstring.reserve(ssize - start - tail); + std::size_t loc = start; + while(loc < ssize - tail) { + // ssize-2 to skip )" at the end + if(escaped_string[loc] == '\\' && (escaped_string[loc + 1] == 'x' || escaped_string[loc + 1] == 'X')) { + auto c1 = escaped_string[loc + 2]; + auto c2 = escaped_string[loc + 3]; + + std::uint32_t res1 = hexConvert(c1); + std::uint32_t res2 = hexConvert(c2); + if(res1 <= 0x0F && res2 <= 0x0F) { + loc += 4; + outstring.push_back(static_cast(res1 * 16 + res2)); + continue; + } + } + outstring.push_back(escaped_string[loc]); + ++loc; + } + return outstring; +} + +CLI11_INLINE void remove_quotes(std::vector &args) { + for(auto &arg : args) { + if(arg.front() == '\"' && arg.back() == '\"') { + remove_quotes(arg); + // only remove escaped for string arguments not literal strings + arg = remove_escaped_characters(arg); + } else { + remove_quotes(arg); + } + } +} + +CLI11_INLINE bool process_quoted_string(std::string &str, char string_char, char literal_char) { + if(str.size() <= 1) { + return false; + } + if(detail::is_binary_escaped_string(str)) { + str = detail::extract_binary_string(str); + return true; + } + if(str.front() == string_char && str.back() == string_char) { + detail::remove_outer(str, string_char); + if(str.find_first_of('\\') != std::string::npos) { + str = detail::remove_escaped_characters(str); + } + return true; + } + if((str.front() == literal_char || str.front() == '`') && str.back() == str.front()) { + detail::remove_outer(str, str.front()); + return true; + } + return false; +} + +std::string get_environment_value(const std::string &env_name) { + char *buffer = nullptr; + std::string ename_string; + +#ifdef _MSC_VER + // Windows version + std::size_t sz = 0; + if(_dupenv_s(&buffer, &sz, env_name.c_str()) == 0 && buffer != nullptr) { + ename_string = std::string(buffer); + free(buffer); + } +#else + // This also works on Windows, but gives a warning + buffer = std::getenv(env_name.c_str()); + if(buffer != nullptr) { + ename_string = std::string(buffer); + } +#endif + return ename_string; } } // namespace detail @@ -722,7 +1337,13 @@ class BadNameString : public ConstructionError { CLI11_ERROR_DEF(ConstructionError, BadNameString) CLI11_ERROR_SIMPLE(BadNameString) static BadNameString OneCharName(std::string name) { return BadNameString("Invalid one char name: " + name); } + static BadNameString MissingDash(std::string name) { + return BadNameString("Long names strings require 2 dashes " + name); + } static BadNameString BadLongName(std::string name) { return BadNameString("Bad long name: " + name); } + static BadNameString BadPositionalName(std::string name) { + return BadNameString("Invalid positional Name: " + name); + } static BadNameString DashesOnly(std::string name) { return BadNameString("Must have a name, not just dashes: " + name); } @@ -1087,11 +1708,19 @@ template class is_direct_constructible { static auto test(int, std::true_type) -> decltype( // NVCC warns about narrowing conversions here #ifdef __CUDACC__ +#ifdef __NVCC_DIAG_PRAGMA_SUPPORT__ +#pragma nv_diag_suppress 2361 +#else #pragma diag_suppress 2361 +#endif #endif TT{std::declval()} #ifdef __CUDACC__ +#ifdef __NVCC_DIAG_PRAGMA_SUPPORT__ +#pragma nv_diag_default 2361 +#else #pragma diag_default 2361 +#endif #endif , std::is_move_assignable()); @@ -1171,8 +1800,10 @@ struct is_mutable_container< decltype(std::declval().clear()), decltype(std::declval().insert(std::declval().end())>(), std::declval()))>, - void>> - : public conditional_t::value, std::false_type, std::true_type> {}; + void>> : public conditional_t::value || + std::is_constructible::value, + std::false_type, + std::true_type> {}; // check to see if an object is a mutable container (fail by default) template struct is_readable_container : std::false_type {}; @@ -1475,6 +2106,8 @@ enum class object_category : int { // string like types string_assignable = 23, string_constructible = 24, + wstring_assignable = 25, + wstring_constructible = 26, other = 45, // special wrapper or container types wrapper_value = 50, @@ -1523,12 +2156,23 @@ template struct classify_object struct classify_object::value>::type> { static constexpr object_category value{object_category::floating_point}; }; +#if defined _MSC_VER +// in MSVC wstring should take precedence if available this isn't as useful on other compilers due to the broader use of +// utf-8 encoding +#define WIDE_STRING_CHECK \ + !std::is_assignable::value && !std::is_constructible::value +#define STRING_CHECK true +#else +#define WIDE_STRING_CHECK true +#define STRING_CHECK !std::is_assignable::value && !std::is_constructible::value +#endif /// String and similar direct assignment template -struct classify_object::value && !std::is_integral::value && - std::is_assignable::value>::type> { +struct classify_object< + T, + typename std::enable_if::value && !std::is_integral::value && WIDE_STRING_CHECK && + std::is_assignable::value>::type> { static constexpr object_category value{object_category::string_assignable}; }; @@ -1538,10 +2182,27 @@ struct classify_object< T, typename std::enable_if::value && !std::is_integral::value && !std::is_assignable::value && (type_count::value == 1) && - std::is_constructible::value>::type> { + WIDE_STRING_CHECK && std::is_constructible::value>::type> { static constexpr object_category value{object_category::string_constructible}; }; +/// Wide strings +template +struct classify_object::value && !std::is_integral::value && + STRING_CHECK && std::is_assignable::value>::type> { + static constexpr object_category value{object_category::wstring_assignable}; +}; + +template +struct classify_object< + T, + typename std::enable_if::value && !std::is_integral::value && + !std::is_assignable::value && (type_count::value == 1) && + STRING_CHECK && std::is_constructible::value>::type> { + static constexpr object_category value{object_category::wstring_constructible}; +}; + /// Enumerations template struct classify_object::value>::type> { static constexpr object_category value{object_category::enumeration}; @@ -1554,12 +2215,13 @@ template struct classify_object struct uncommon_type { - using type = typename std::conditional::value && !std::is_integral::value && - !std::is_assignable::value && - !std::is_constructible::value && !is_complex::value && - !is_mutable_container::value && !std::is_enum::value, - std::true_type, - std::false_type>::type; + using type = typename std::conditional< + !std::is_floating_point::value && !std::is_integral::value && + !std::is_assignable::value && !std::is_constructible::value && + !std::is_assignable::value && !std::is_constructible::value && + !is_complex::value && !is_mutable_container::value && !std::is_enum::value, + std::true_type, + std::false_type>::type; static constexpr bool value = type::value; }; @@ -1748,7 +2410,7 @@ bool integral_conversion(const std::string &input, T &output) noexcept { if(input.empty() || input.front() == '-') { return false; } - char *val = nullptr; + char *val{nullptr}; errno = 0; std::uint64_t output_ll = std::strtoull(input.c_str(), &val, 0); if(errno == ERANGE) { @@ -1764,6 +2426,33 @@ bool integral_conversion(const std::string &input, T &output) noexcept { output = (output_sll < 0) ? static_cast(0) : static_cast(output_sll); return (static_cast(output) == output_sll); } + // remove separators + if(input.find_first_of("_'") != std::string::npos) { + std::string nstring = input; + nstring.erase(std::remove(nstring.begin(), nstring.end(), '_'), nstring.end()); + nstring.erase(std::remove(nstring.begin(), nstring.end(), '\''), nstring.end()); + return integral_conversion(nstring, output); + } + if(input.compare(0, 2, "0o") == 0) { + val = nullptr; + errno = 0; + output_ll = std::strtoull(input.c_str() + 2, &val, 8); + if(errno == ERANGE) { + return false; + } + output = static_cast(output_ll); + return (val == (input.c_str() + input.size()) && static_cast(output) == output_ll); + } + if(input.compare(0, 2, "0b") == 0) { + val = nullptr; + errno = 0; + output_ll = std::strtoull(input.c_str() + 2, &val, 2); + if(errno == ERANGE) { + return false; + } + output = static_cast(output_ll); + return (val == (input.c_str() + input.size()) && static_cast(output) == output_ll); + } return false; } @@ -1788,11 +2477,38 @@ bool integral_conversion(const std::string &input, T &output) noexcept { output = static_cast(1); return true; } + // remove separators + if(input.find_first_of("_'") != std::string::npos) { + std::string nstring = input; + nstring.erase(std::remove(nstring.begin(), nstring.end(), '_'), nstring.end()); + nstring.erase(std::remove(nstring.begin(), nstring.end(), '\''), nstring.end()); + return integral_conversion(nstring, output); + } + if(input.compare(0, 2, "0o") == 0) { + val = nullptr; + errno = 0; + output_ll = std::strtoll(input.c_str() + 2, &val, 8); + if(errno == ERANGE) { + return false; + } + output = static_cast(output_ll); + return (val == (input.c_str() + input.size()) && static_cast(output) == output_ll); + } + if(input.compare(0, 2, "0b") == 0) { + val = nullptr; + errno = 0; + output_ll = std::strtoll(input.c_str() + 2, &val, 2); + if(errno == ERANGE) { + return false; + } + output = static_cast(output_ll); + return (val == (input.c_str() + input.size()) && static_cast(output) == output_ll); + } return false; } -/// Convert a flag into an integer value typically binary flags -inline std::int64_t to_flag_value(std::string val) { +/// Convert a flag into an integer value typically binary flags sets errno to nonzero if conversion failed +inline std::int64_t to_flag_value(std::string val) noexcept { static const std::string trueString("true"); static const std::string falseString("false"); if(val == trueString) { @@ -1820,7 +2536,8 @@ inline std::int64_t to_flag_value(std::string val) { ret = 1; break; default: - throw std::invalid_argument("unrecognized character"); + errno = EINVAL; + return -1; } return ret; } @@ -1829,7 +2546,11 @@ inline std::int64_t to_flag_value(std::string val) { } else if(val == falseString || val == "off" || val == "no" || val == "disable") { ret = -1; } else { - ret = std::stoll(val); + char *loc_ptr{nullptr}; + ret = std::strtoll(val.c_str(), &loc_ptr, 0); + if(loc_ptr != (val.c_str() + val.size()) && errno == 0) { + errno = EINVAL; + } } return ret; } @@ -1858,18 +2579,16 @@ bool lexical_cast(const std::string &input, T &output) { template ::value == object_category::boolean_value, detail::enabler> = detail::dummy> bool lexical_cast(const std::string &input, T &output) { - try { - auto out = to_flag_value(input); + errno = 0; + auto out = to_flag_value(input); + if(errno == 0) { output = (out > 0); - return true; - } catch(const std::invalid_argument &) { - return false; - } catch(const std::out_of_range &) { - // if the number is out of the range of a 64 bit value then it is still a number and for this purpose is still - // valid all we care about the sign + } else if(errno == ERANGE) { output = (input[0] != '-'); - return true; + } else { + return false; } + return true; } /// Floats @@ -1882,7 +2601,17 @@ bool lexical_cast(const std::string &input, T &output) { char *val = nullptr; auto output_ld = std::strtold(input.c_str(), &val); output = static_cast(output_ld); - return val == (input.c_str() + input.size()); + if(val == (input.c_str() + input.size())) { + return true; + } + // remove separators + if(input.find_first_of("_'") != std::string::npos) { + std::string nstring = input; + nstring.erase(std::remove(nstring.begin(), nstring.end(), '_'), nstring.end()); + nstring.erase(std::remove(nstring.begin(), nstring.end(), '\''), nstring.end()); + return lexical_cast(nstring, output); + } + return false; } /// complex @@ -1934,6 +2663,23 @@ bool lexical_cast(const std::string &input, T &output) { return true; } +/// Wide strings +template < + typename T, + enable_if_t::value == object_category::wstring_assignable, detail::enabler> = detail::dummy> +bool lexical_cast(const std::string &input, T &output) { + output = widen(input); + return true; +} + +template < + typename T, + enable_if_t::value == object_category::wstring_constructible, detail::enabler> = detail::dummy> +bool lexical_cast(const std::string &input, T &output) { + output = T{widen(input)}; + return true; +} + /// Enumerations template ::value == object_category::enumeration, detail::enabler> = detail::dummy> @@ -2062,7 +2808,9 @@ template ::value && (classify_object::value == object_category::string_assignable || - classify_object::value == object_category::string_constructible), + classify_object::value == object_category::string_constructible || + classify_object::value == object_category::wstring_assignable || + classify_object::value == object_category::wstring_constructible), detail::enabler> = detail::dummy> bool lexical_assign(const std::string &input, AssignTo &output) { return lexical_cast(input, output); @@ -2073,7 +2821,9 @@ template ::value && std::is_assignable::value && classify_object::value != object_category::string_assignable && - classify_object::value != object_category::string_constructible, + classify_object::value != object_category::string_constructible && + classify_object::value != object_category::wstring_assignable && + classify_object::value != object_category::wstring_constructible, detail::enabler> = detail::dummy> bool lexical_assign(const std::string &input, AssignTo &output) { if(input.empty()) { @@ -2112,9 +2862,17 @@ bool lexical_assign(const std::string &input, AssignTo &output) { output = 0; return true; } - int val = 0; + int val{0}; if(lexical_cast(input, val)) { +#if defined(__clang__) +/* on some older clang compilers */ +#pragma clang diagnostic push +#pragma clang diagnostic ignored "-Wsign-conversion" +#endif output = val; +#if defined(__clang__) +#pragma clang diagnostic pop +#endif return true; } return false; @@ -2169,12 +2927,12 @@ template = detail::dummy> bool lexical_conversion(const std::vector &strings, AssignTo &output) { // the remove const is to handle pair types coming from a container - typename std::remove_const::type>::type v1; - typename std::tuple_element<1, ConvertTo>::type v2; - bool retval = lexical_assign(strings[0], v1); - if(strings.size() > 1) { - retval = retval && lexical_assign(strings[1], v2); - } + using FirstType = typename std::remove_const::type>::type; + using SecondType = typename std::tuple_element<1, ConvertTo>::type; + FirstType v1; + SecondType v2; + bool retval = lexical_assign(strings[0], v1); + retval = retval && lexical_assign((strings.size() > 1) ? strings[1] : std::string{}, v2); if(retval) { output = AssignTo{v1, v2}; } @@ -2189,6 +2947,9 @@ template = detail::dummy> bool lexical_conversion(const std::vector &strings, AssignTo &output) { output.erase(output.begin(), output.end()); + if(strings.empty()) { + return true; + } if(strings.size() == 1 && strings[0] == "{}") { return true; } @@ -2333,7 +3094,7 @@ tuple_type_conversion(std::vector &strings, AssignTo &output) { std::size_t index{subtype_count_min::value}; const std::size_t mx_count{subtype_count::value}; - const std::size_t mx{(std::max)(mx_count, strings.size())}; + const std::size_t mx{(std::min)(mx_count, strings.size() - 1)}; while(index < mx) { if(is_separator(strings[index])) { @@ -2343,7 +3104,11 @@ tuple_type_conversion(std::vector &strings, AssignTo &output) { } bool retval = lexical_conversion( std::vector(strings.begin(), strings.begin() + static_cast(index)), output); - strings.erase(strings.begin(), strings.begin() + static_cast(index) + 1); + if(strings.size() > index) { + strings.erase(strings.begin(), strings.begin() + static_cast(index) + 1); + } else { + strings.clear(); + } return retval; } @@ -2487,12 +3252,13 @@ inline std::string sum_string_vector(const std::vector &values) { double tv{0.0}; auto comp = lexical_cast(arg, tv); if(!comp) { - try { - tv = static_cast(detail::to_flag_value(arg)); - } catch(const std::exception &) { - fail = true; + errno = 0; + auto fv = detail::to_flag_value(arg); + fail = (errno != 0); + if(fail) { break; } + tv = static_cast(fv); } val += tv; } @@ -2501,13 +3267,10 @@ inline std::string sum_string_vector(const std::vector &values) { output.append(arg); } } else { - if(val <= static_cast((std::numeric_limits::min)()) || - val >= static_cast((std::numeric_limits::max)()) || - std::ceil(val) == std::floor(val)) { - output = detail::value_string(static_cast(val)); - } else { - output = detail::value_string(val); - } + std::ostringstream out; + out.precision(16); + out << val; + output = out.str(); } return output; } @@ -2553,7 +3316,7 @@ CLI11_INLINE bool split_short(const std::string ¤t, std::string &name, std } CLI11_INLINE bool split_long(const std::string ¤t, std::string &name, std::string &value) { - if(current.size() > 2 && current.substr(0, 2) == "--" && valid_first_char(current[2])) { + if(current.size() > 2 && current.compare(0, 2, "--") == 0 && valid_first_char(current[2])) { auto loc = current.find_first_of('='); if(loc != std::string::npos) { name = current.substr(2, loc - 2); @@ -2625,7 +3388,6 @@ get_names(const std::vector &input) { std::vector short_names; std::vector long_names; std::string pos_name; - for(std::string name : input) { if(name.length() == 0) { continue; @@ -2633,6 +3395,8 @@ get_names(const std::vector &input) { if(name.length() > 1 && name[0] == '-' && name[1] != '-') { if(name.length() == 2 && valid_first_char(name[1])) short_names.emplace_back(1, name[1]); + else if(name.length() > 2) + throw BadNameString::MissingDash(name); else throw BadNameString::OneCharName(name); } else if(name.length() > 2 && name.substr(0, 2) == "--") { @@ -2644,12 +3408,15 @@ get_names(const std::vector &input) { } else if(name == "-" || name == "--") { throw BadNameString::DashesOnly(name); } else { - if(pos_name.length() > 0) + if(!pos_name.empty()) throw BadNameString::MultiPositionalNames(name); - pos_name = name; + if(valid_name_string(name)) { + pos_name = name; + } else { + throw BadNameString::BadPositionalName(name); + } } } - return std::make_tuple(short_names, long_names, pos_name); } @@ -2666,7 +3433,6 @@ struct ConfigItem { /// This is the name std::string name{}; - /// Listing of inputs std::vector inputs{}; @@ -2729,8 +3495,8 @@ class ConfigBase : public Config { char valueDelimiter = '='; /// the character to use around strings char stringQuote = '"'; - /// the character to use around single characters - char characterQuote = '\''; + /// the character to use around single characters and literal strings + char literalQuote = '\''; /// the maximum number of layers to allow uint8_t maximumLayers{255}; /// the separator used to separator parent layers @@ -2766,10 +3532,10 @@ class ConfigBase : public Config { valueDelimiter = vSep; return this; } - /// Specify the quote characters used around strings and characters - ConfigBase *quoteCharacter(char qString, char qChar) { + /// Specify the quote characters used around strings and literal strings + ConfigBase *quoteCharacter(char qString, char literalChar) { stringQuote = qString; - characterQuote = qChar; + literalQuote = literalChar; return this; } /// Specify the maximum number of parents @@ -3001,6 +3767,11 @@ class IPV4Validator : public Validator { IPV4Validator(); }; +class EscapedStringTransformer : public Validator { + public: + EscapedStringTransformer(); +}; + } // namespace detail // Static is not needed here, because global const implies static. @@ -3020,6 +3791,9 @@ const detail::NonexistentPathValidator NonexistentPath; /// Check for an IP4 address const detail::IPV4Validator ValidIPV4; +/// convert escaped characters into their associated values +const detail::EscapedStringTransformer EscapedString; + /// Validate the input as a particular type template class TypeValidator : public Validator { public: @@ -3772,14 +4546,14 @@ namespace detail { #if defined CLI11_HAS_FILESYSTEM && CLI11_HAS_FILESYSTEM > 0 CLI11_INLINE path_type check_path(const char *file) noexcept { std::error_code ec; - auto stat = std::filesystem::status(file, ec); + auto stat = std::filesystem::status(to_path(file), ec); if(ec) { return path_type::nonexistent; } switch(stat.type()) { case std::filesystem::file_type::none: // LCOV_EXCL_LINE case std::filesystem::file_type::not_found: - return path_type::nonexistent; + return path_type::nonexistent; // LCOV_EXCL_LINE case std::filesystem::file_type::directory: return path_type::directory; case std::filesystem::file_type::symlink: @@ -3873,10 +4647,29 @@ CLI11_INLINE IPV4Validator::IPV4Validator() : Validator("IPV4") { return std::string("Each IP number must be between 0 and 255 ") + var; } } - return std::string(); + return std::string{}; }; } +CLI11_INLINE EscapedStringTransformer::EscapedStringTransformer() { + func_ = [](std::string &str) { + try { + if(str.size() > 1 && (str.front() == '\"' || str.front() == '\'' || str.front() == '`') && + str.front() == str.back()) { + process_quoted_string(str); + } else if(str.find_first_of('\\') != std::string::npos) { + if(detail::is_binary_escaped_string(str)) { + str = detail::extract_binary_string(str); + } else { + str = remove_escaped_characters(str); + } + } + return std::string{}; + } catch(const std::invalid_argument &ia) { + return std::string(ia.what()); + } + }; +} } // namespace detail CLI11_INLINE FileOnDefaultPath::FileOnDefaultPath(std::string default_path, bool enableErrorReturn) @@ -4175,7 +4968,8 @@ enum class MultiOptionPolicy : char { TakeFirst, //!< take only the first Expected number of arguments Join, //!< merge all the arguments together into a single string via the delimiter character default('\n') TakeAll, //!< just get all the passed argument regardless - Sum //!< sum all the arguments together if numerical or concatenate directly without delimiter + Sum, //!< sum all the arguments together if numerical or concatenate directly without delimiter + Reverse, //!< take only the last Expected number of arguments in reverse order }; /// This is the CRTP base class for Option and OptionDefaults. It was designed this way @@ -4683,12 +5477,12 @@ class Option : public OptionBase