This library is considered feature-complete by its creator. It will still receive bug fixes, and good pull requests will be accepted, but no new major functionality or API changes will be added to this codebase.
A simple, small, flexible, single-header C++11 argument parsing library.
This is designed to appear somewhat similar to Python's argparse, but in C++, with static type checking, and hopefully a lot faster (also allowing fully nestable group logic, where Python's argparse does not).
UTF-8 support is limited at best. No normalization is performed, so non-ascii characters are very best kept out of flags, and combined glyphs are probably going to mess up help output if you use them. Most UTF-8 necessary for internationalization should work for most cases, though heavily combinatory UTF alphabets may wreak havoc.
This program is MIT-licensed, so you can use the header as-is with no restrictions. I'd appreciate attribution in a README, Man page, or something if you are feeling generous, but all that's required is that you don't remove the license and my name from the header of the args.hxx file in source redistributions (ie. don't pretend that you wrote it). I do welcome additions and updates wherever you feel like contributing code.
The API documentation can be found at https://taywee.github.io/args
The code can be downloaded at https://github.com/Taywee/args
There are also somewhat extensive examples below.
You can find the complete test cases at https://github.com/Taywee/args/blob/master/test.cxx, which should very well describe the usage, as it's built to push the boundaries.
It:
- Lets you handle flags, flag+value, and positional arguments simply and elegantly, with the full help of static typechecking.
- Allows you to use your own types in a pretty simple way.
- Lets you use count flags, and lists of all argument-accepting types.
- Allows full validation of groups of required arguments, though output isn't pretty when something fails group validation. User validation functions are accepted. Groups are fully nestable.
- Generates pretty help for you, with some good tweakable parameters.
- Lets you customize all prefixes and most separators, allowing you to create an infinite number of different argument syntaxes.
- Lets you parse, by default, any type that has a stream extractor operator for it. If this doesn't work for your uses, you can supply a function and parse the string yourself if you like.
- Lets you decide not to allow separate-argument value flags or joined ones
(like disallowing
--foo bar
, requiring--foo=bar
, or the inverse, or the same for short options). - Allows you to create subparsers, to reuse arguments for multiple commands and to refactor your command's logic to a function or lambda.
- Allows one value flag to take a specific number of values (like
--foo first second
, where --foo slurps both arguments). - Allows you to have value flags only optionally accept values.
- Provides autocompletion for bash.
There are tons of things this library does not do!
- Allow you to intermix multiple different prefix types (eg.
++foo
and--foo
in the same parser), though shortopt and longopt prefixes can be different. - Allow you to make flags sensitive to order (like gnu find), or make them
sensitive to relative ordering with positionals. The only orderings that are
order-sensitive are:
- Positionals relative to one-another
- List positionals or flag values to each of their own respective items
- Allow you to use a positional list before any other positionals (the last argument list will slurp all subsequent positional arguments). The logic for allowing this would be a lot more code than I'd like, and would make static checking much more difficult, requiring us to sort std::string arguments and pair them to positional arguments before assigning them, rather than what we currently do, which is assigning them as we go for better simplicity and speed. The library doesn't stop you from trying, but the first positional list will slurp in all following positionals
sudo make install
Or, to install it somewhere special (default is /usr/local
):
sudo make install DESTDIR=/opt/mydir
You can also copy the file into your source tree, if you want to be absolutely sure you keep a stable API between projects.
If you prefer other installation methods, many standard ones are available and included, including CMake, conan, buck, and meson. An example CMake file using args is included in the examples directory.
make doc/man
sudo make installman
This requires Doxygen
doxygen Doxyfile
Your docs are now in doc/html
Simply add the following entry to your .tipi/deps
file
{
"taywee/args": { "@": "6.4.1" }
}
You can optionally remove the @
section to target HEAD easily.
Create an ArgumentParser, modify its attributes to fit your needs, add arguments through regular argument objects (or create your own), and match them with an args::Matcher object (check its construction details in the doxygen documentation.
Then you can either call it with args::ArgumentParser::ParseCLI for the full command line with program name, or args::ArgumentParser::ParseArgs with just the arguments to be parsed. The argument and group variables can then be interpreted as a boolean to see if they've been matched.
All variables can be pulled (including the boolean match status for regular args::Flag variables) with args::get.
This is unfortunately not possible, given the power of the groups available.
For instance, if you have a group validation that works like
(A && B) || (C && (D XOR E))
, how is this library going to be able to
determine what exactly when wrong when it fails? It only knows that the
entire expression evaluated false, not specifically what the user did wrong
(and this is doubled over by the fact that validation operations are ordinary
functions without any special meaning to the library). As you are the only one
who understands the logic of your program, if you want useful group messages,
you have to catch the ValidationError as a special case and check your own
groups and spit out messages accordingly.
Yes. tests.cxx in the git repository has a set of standard tests (which are still relatively small in number, but I would welcome some expansion here), and thanks to Travis CI and AppVeyor, these tests run with every single push:
% make runtests
g++ test.cxx -o test.o -I. -std=c++11 -O2 -c -MMD
g++ -o test test.o -std=c++11 -O2
./test
===============================================================================
All tests passed (74 assertions in 15 test cases)
%
The testing library used is Catch.
All the code examples here will be complete code examples, with some output.
#include <iostream>
#include <args.hxx>
int main(int argc, char **argv)
{
args::ArgumentParser parser("This is a test program.", "This goes after the options.");
args::HelpFlag help(parser, "help", "Display this help menu", {'h', "help"});
args::CompletionFlag completion(parser, {"complete"});
try
{
parser.ParseCLI(argc, argv);
}
catch (const args::Completion& e)
{
std::cout << e.what();
return 0;
}
catch (const args::Help&)
{
std::cout << parser;
return 0;
}
catch (const args::ParseError& e)
{
std::cerr << e.what() << std::endl;
std::cerr << parser;
return 1;
}
return 0;
}
% ./test
% ./test -h
./test {OPTIONS}
This is a test program.
OPTIONS:
-h, --help Display this help menu
This goes after the options.
%
#include <iostream>
#include <args.hxx>
int main(int argc, char **argv)
{
args::ArgumentParser parser("This is a test program.", "This goes after the options.");
args::Group group(parser, "This group is all exclusive:", args::Group::Validators::Xor);
args::Flag foo(group, "foo", "The foo flag", {'f', "foo"});
args::Flag bar(group, "bar", "The bar flag", {'b'});
args::Flag baz(group, "baz", "The baz flag", {"baz"});
try
{
parser.ParseCLI(argc, argv);
}
catch (args::Help)
{
std::cout << parser;
return 0;
}
catch (args::ParseError e)
{
std::cerr << e.what() << std::endl;
std::cerr << parser;
return 1;
}
catch (args::ValidationError e)
{
std::cerr << e.what() << std::endl;
std::cerr << parser;
return 1;
}
if (foo) { std::cout << "foo" << std::endl; }
if (bar) { std::cout << "bar" << std::endl; }
if (baz) { std::cout << "baz" << std::endl; }
return 0;
}
% ./test
Group validation failed somewhere!
./test {OPTIONS}
This is a test program.
OPTIONS:
This group is all exclusive:
-f, --foo The foo flag
-b The bar flag
--baz The baz flag
This goes after the options.
% ./test -f
foo
% ./test --foo
foo
% ./test --foo -f
foo
% ./test -b
bar
% ./test --baz
baz
% ./test --baz -f
Group validation failed somewhere!
./test {OPTIONS}
This is a test program.
...
% ./test --baz -fb
Group validation failed somewhere!
./test {OPTIONS}
...
%
#include <iostream>
#include <args.hxx>
int main(int argc, char **argv)
{
args::ArgumentParser parser("This is a test program.", "This goes after the options.");
args::HelpFlag help(parser, "help", "Display this help menu", {'h', "help"});
args::ValueFlag<int> integer(parser, "integer", "The integer flag", {'i'});
args::ValueFlagList<char> characters(parser, "characters", "The character flag", {'c'});
args::Positional<std::string> foo(parser, "foo", "The foo position");
args::PositionalList<double> numbers(parser, "numbers", "The numbers position list");
try
{
parser.ParseCLI(argc, argv);
}
catch (args::Help)
{
std::cout << parser;
return 0;
}
catch (args::ParseError e)
{
std::cerr << e.what() << std::endl;
std::cerr << parser;
return 1;
}
catch (args::ValidationError e)
{
std::cerr << e.what() << std::endl;
std::cerr << parser;
return 1;
}
if (integer) { std::cout << "i: " << args::get(integer) << std::endl; }
if (characters) { for (const auto ch: args::get(characters)) { std::cout << "c: " << ch << std::endl; } }
if (foo) { std::cout << "f: " << args::get(foo) << std::endl; }
if (numbers) { for (const auto nm: args::get(numbers)) { std::cout << "n: " << nm << std::endl; } }
return 0;
}
% ./test -h
./test {OPTIONS} [foo] [numbers...]
This is a test program.
OPTIONS:
-h, --help Display this help menu
-i integer The integer flag
-c characters The character flag
foo The foo position
numbers The numbers position list
"--" can be used to terminate flag options and force all following
arguments to be treated as positional options
This goes after the options.
% ./test -i 5
i: 5
% ./test -i 5.2
Argument 'integer' received invalid value type '5.2'
./test {OPTIONS} [foo] [numbers...]
% ./test -c 1 -c 2 -c 3
c: 1
c: 2
c: 3
%
% ./test 1 2 3 4 5 6 7 8 9
f: 1
n: 2
n: 3
n: 4
n: 5
n: 6
n: 7
n: 8
n: 9
% ./test 1 2 3 4 5 6 7 8 9 a
Argument 'numbers' received invalid value type 'a'
./test {OPTIONS} [foo] [numbers...]
This is a test program.
...
#include <iostream>
#include <args.hxx>
int main(int argc, char **argv)
{
args::ArgumentParser p("git-like parser");
args::Group commands(p, "commands");
args::Command add(commands, "add", "add file contents to the index");
args::Command commit(commands, "commit", "record changes to the repository");
args::Group arguments(p, "arguments", args::Group::Validators::DontCare, args::Options::Global);
args::ValueFlag<std::string> gitdir(arguments, "path", "", {"git-dir"});
args::HelpFlag h(arguments, "help", "help", {'h', "help"});
args::PositionalList<std::string> pathsList(arguments, "paths", "files to commit");
try
{
p.ParseCLI(argc, argv);
if (add)
{
std::cout << "Add";
}
else
{
std::cout << "Commit";
}
for (auto &&path : pathsList)
{
std::cout << ' ' << path;
}
std::cout << std::endl;
}
catch (args::Help)
{
std::cout << p;
}
catch (args::Error& e)
{
std::cerr << e.what() << std::endl << p;
return 1;
}
return 0;
}
% ./test -h
./test COMMAND [paths...] {OPTIONS}
git-like parser
OPTIONS:
commands
add add file contents to the index
commit record changes to the repository
arguments
--git-dir=[path]
-h, --help help
paths... files
"--" can be used to terminate flag options and force all following
arguments to be treated as positional options
% ./test add 1 2
Add 1 2
#include <iostream>
#include "args.hxx"
args::Group arguments("arguments");
args::ValueFlag<std::string> gitdir(arguments, "path", "", {"git-dir"});
args::HelpFlag h(arguments, "help", "help", {'h', "help"});
args::PositionalList<std::string> pathsList(arguments, "paths", "files to commit");
void CommitCommand(args::Subparser &parser)
{
args::ValueFlag<std::string> message(parser, "MESSAGE", "commit message", {'m'});
parser.Parse();
std::cout << "Commit";
for (auto &&path : pathsList)
{
std::cout << ' ' << path;
}
std::cout << std::endl;
if (message)
{
std::cout << "message: " << args::get(message) << std::endl;
}
}
int main(int argc, const char **argv)
{
args::ArgumentParser p("git-like parser");
args::Group commands(p, "commands");
args::Command add(commands, "add", "add file contents to the index", [&](args::Subparser &parser)
{
parser.Parse();
std::cout << "Add";
for (auto &&path : pathsList)
{
std::cout << ' ' << path;
}
std::cout << std::endl;
});
args::Command commit(commands, "commit", "record changes to the repository", &CommitCommand);
args::GlobalOptions globals(p, arguments);
try
{
p.ParseCLI(argc, argv);
}
catch (args::Help)
{
std::cout << p;
}
catch (args::Error& e)
{
std::cerr << e.what() << std::endl << p;
return 1;
}
return 0;
}
% ./test -h
./test COMMAND [paths...] {OPTIONS}
git-like parser
OPTIONS:
commands
add add file contents to the index
commit record changes to the repository
arguments
--git-dir=[path]
-h, --help help
paths... files
"--" can be used to terminate flag options and force all following
arguments to be treated as positional options
% ./test add 1 2
Add 1 2
% ./test commit -m "my commit message" 1 2
Commit 1 2
message: my commit message
#include <iostream>
#include <tuple>
std::istream& operator>>(std::istream& is, std::tuple<int, int>& ints)
{
is >> std::get<0>(ints);
is.get();
is >> std::get<1>(ints);
return is;
}
#include <args.hxx>
struct DoublesReader
{
void operator()(const std::string &name, const std::string &value, std::tuple<double, double> &destination)
{
size_t commapos = 0;
std::get<0>(destination) = std::stod(value, &commapos);
std::get<1>(destination) = std::stod(std::string(value, commapos + 1));
}
};
int main(int argc, char **argv)
{
args::ArgumentParser parser("This is a test program.");
args::Positional<std::tuple<int, int>> ints(parser, "INTS", "This takes a pair of integers.");
args::Positional<std::tuple<double, double>, DoublesReader> doubles(parser, "DOUBLES", "This takes a pair of doubles.");
try
{
parser.ParseCLI(argc, argv);
}
catch (args::Help)
{
std::cout << parser;
return 0;
}
catch (args::ParseError e)
{
std::cerr << e.what() << std::endl;
std::cerr << parser;
return 1;
}
if (ints)
{
std::cout << "ints found: " << std::get<0>(args::get(ints)) << " and " << std::get<1>(args::get(ints)) << std::endl;
}
if (doubles)
{
std::cout << "doubles found: " << std::get<0>(args::get(doubles)) << " and " << std::get<1>(args::get(doubles)) << std::endl;
}
return 0;
}
% ./test -h
Argument could not be matched: 'h'
./test [INTS] [DOUBLES]
This is a test program.
OPTIONS:
INTS This takes a pair of integers.
DOUBLES This takes a pair of doubles.
% ./test 5
ints found: 5 and 0
% ./test 5,8
ints found: 5 and 8
% ./test 5,8 2.4,8
ints found: 5 and 8
doubles found: 2.4 and 8
% ./test 5,8 2.4,
terminate called after throwing an instance of 'std::invalid_argument'
what(): stod
zsh: abort ./test 5,8 2.4,
% ./test 5,8 2.4
terminate called after throwing an instance of 'std::out_of_range'
what(): basic_string::basic_string: __pos (which is 4) > this->size() (which is 3)
zsh: abort ./test 5,8 2.4
% ./test 5,8 2.4-7
ints found: 5 and 8
doubles found: 2.4 and 7
% ./test 5,8 2.4,-7
ints found: 5 and 8
doubles found: 2.4 and -7
As you can see, with your own types, validation can get a little weird. Make sure to check and throw a parsing error (or whatever error you want to catch) if you can't fully deduce your type. The built-in validator will only throw if there are unextracted characters left in the stream.
#include <iostream>
#include <args.hxx>
int main(int argc, char **argv)
{
args::ArgumentParser parser("This is a test program with a really long description that is probably going to have to be wrapped across multiple different lines. This is a test to see how the line wrapping works", "This goes after the options. This epilog is also long enough that it will have to be properly wrapped to display correctly on the screen");
args::HelpFlag help(parser, "HELP", "Show this help menu.", {'h', "help"});
args::ValueFlag<std::string> foo(parser, "FOO", "The foo flag.", {'a', 'b', 'c', "a", "b", "c", "the-foo-flag"});
args::ValueFlag<std::string> bar(parser, "BAR", "The bar flag. This one has a lot of options, and will need wrapping in the description, along with its long flag list.", {'d', 'e', 'f', "d", "e", "f"});
args::ValueFlag<std::string> baz(parser, "FOO", "The baz flag. This one has a lot of options, and will need wrapping in the description, even with its short flag list.", {"baz"});
args::Positional<std::string> pos1(parser, "POS1", "The pos1 argument.");
args::PositionalList<std::string> poslist1(parser, "POSLIST1", "The poslist1 argument.");
args::Positional<std::string> pos2(parser, "POS2", "The pos2 argument.");
args::PositionalList<std::string> poslist2(parser, "POSLIST2", "The poslist2 argument.");
args::Positional<std::string> pos3(parser, "POS3", "The pos3 argument.");
args::PositionalList<std::string> poslist3(parser, "POSLIST3", "The poslist3 argument.");
args::Positional<std::string> pos4(parser, "POS4", "The pos4 argument.");
args::PositionalList<std::string> poslist4(parser, "POSLIST4", "The poslist4 argument.");
args::Positional<std::string> pos5(parser, "POS5", "The pos5 argument.");
args::PositionalList<std::string> poslist5(parser, "POSLIST5", "The poslist5 argument.");
args::Positional<std::string> pos6(parser, "POS6", "The pos6 argument.");
args::PositionalList<std::string> poslist6(parser, "POSLIST6", "The poslist6 argument.");
args::Positional<std::string> pos7(parser, "POS7", "The pos7 argument.");
args::PositionalList<std::string> poslist7(parser, "POSLIST7", "The poslist7 argument.");
args::Positional<std::string> pos8(parser, "POS8", "The pos8 argument.");
args::PositionalList<std::string> poslist8(parser, "POSLIST8", "The poslist8 argument.");
args::Positional<std::string> pos9(parser, "POS9", "The pos9 argument.");
args::PositionalList<std::string> poslist9(parser, "POSLIST9", "The poslist9 argument.");
try
{
parser.ParseCLI(argc, argv);
}
catch (args::Help)
{
std::cout << parser;
return 0;
}
catch (args::ParseError e)
{
std::cerr << e.what() << std::endl;
std::cerr << parser;
return 1;
}
catch (args::ValidationError e)
{
std::cerr << e.what() << std::endl;
std::cerr << parser;
return 1;
}
return 0;
}
% ./test -h
./test {OPTIONS} [POS1] [POSLIST1...] [POS2] [POSLIST2...] [POS3]
[POSLIST3...] [POS4] [POSLIST4...] [POS5] [POSLIST5...] [POS6]
[POSLIST6...] [POS7] [POSLIST7...] [POS8] [POSLIST8...] [POS9]
[POSLIST9...]
This is a test program with a really long description that is probably going
to have to be wrapped across multiple different lines. This is a test to see
how the line wrapping works
OPTIONS:
-h, --help Show this help menu.
-a FOO, -b FOO, -c FOO, --a FOO, --b FOO, --c FOO, --the-foo-flag FOO
The foo flag.
-d BAR, -e BAR, -f BAR, --d BAR, --e BAR, --f BAR
The bar flag. This one has a lot of options, and will
need wrapping in the description, along with its long
flag list.
--baz FOO The baz flag. This one has a lot of options, and will
need wrapping in the description, even with its short
flag list.
POS1 The pos1 argument.
POSLIST1 The poslist1 argument.
POS2 The pos2 argument.
POSLIST2 The poslist2 argument.
POS3 The pos3 argument.
POSLIST3 The poslist3 argument.
POS4 The pos4 argument.
POSLIST4 The poslist4 argument.
POS5 The pos5 argument.
POSLIST5 The poslist5 argument.
POS6 The pos6 argument.
POSLIST6 The poslist6 argument.
POS7 The pos7 argument.
POSLIST7 The poslist7 argument.
POS8 The pos8 argument.
POSLIST8 The poslist8 argument.
POS9 The pos9 argument.
POSLIST9 The poslist9 argument.
"--" can be used to terminate flag options and force all following
arguments to be treated as positional options
This goes after the options. This epilog is also long enough that it will
have to be properly wrapped to display correctly on the screen
%
Completion currently only supports bash.
#include <iostream>
#include <args.hxx>
int main(int argc, char **argv)
{
args::ArgumentParser parser("This is a test program.", "This goes after the options.");
args::HelpFlag help(parser, "help", "Display this help menu", {'h', "help"});
args::CompletionFlag completion(parser, {"complete"});
args::ValueFlag<std::string> foo(parser, "name", "description", {'f', "foo"}, "abc");
args::ValueFlag<std::string> bar(parser, "name", "description", {'b', "bar"}, "abc");
args::ValueFlag<std::string> baz(parser, "name", "description", {'a', "baz"}, "abc");
try
{
parser.ParseCLI(argc, argv);
}
catch (const args::Completion& e)
{
std::cout << e.what();
return 0;
}
catch (const args::Help&)
{
std::cout << parser;
return 0;
}
catch (const args::ParseError& e)
{
std::cerr << e.what() << std::endl;
std::cerr << parser;
return 1;
}
return 0;
}
g++ -ocompletiontest ./completiontest.cxx "-I$args_path"
PATH="$PWD:$PATH"
_args() {
_init_completion -n 2> /dev/null
local program comparg
program="${COMP_WORDS[0]}"
comparg="--complete" # replace this with your flag
COMPREPLY=($("$program" "$comparg" bash "$COMP_CWORD" "${COMP_WORDS[@]}" 2> /dev/null))
[[ $COMPREPLY ]] && return
_filedir
}
complete -F _args completiontest
# Now when you type `completiontest --` and press tab, you'll get completion
#include <iostream>
#include <args.hxx>
int main(int argc, char **argv)
{
args::ArgumentParser parser("This command likes to break your disks");
parser.LongPrefix("");
parser.LongSeparator("=");
args::HelpFlag help(parser, "HELP", "Show this help menu.", {"help"});
args::ValueFlag<long> bs(parser, "BYTES", "Block size", {"bs"}, 512);
args::ValueFlag<long> skip(parser, "BYTES", "Bytes to skip", {"skip"}, 0);
args::ValueFlag<std::string> input(parser, "BLOCK SIZE", "Block size", {"if"});
args::ValueFlag<std::string> output(parser, "BLOCK SIZE", "Block size", {"of"});
try
{
parser.ParseCLI(argc, argv);
}
catch (args::Help)
{
std::cout << parser;
return 0;
}
catch (args::ParseError e)
{
std::cerr << e.what() << std::endl;
std::cerr << parser;
return 1;
}
catch (args::ValidationError e)
{
std::cerr << e.what() << std::endl;
std::cerr << parser;
return 1;
}
std::cout << "bs = " << args::get(bs) << std::endl;
std::cout << "skip = " << args::get(skip) << std::endl;
if (input) { std::cout << "if = " << args::get(input) << std::endl; }
if (output) { std::cout << "of = " << args::get(output) << std::endl; }
return 0;
}
% ./test help
./test {OPTIONS}
This command likes to break your disks
OPTIONS:
help Show this help menu.
bs=[BYTES] Block size
skip=[BYTES] Bytes to skip
if=[BLOCK SIZE] Block size
of=[BLOCK SIZE] Block size
% ./test bs=1024 skip=7 if=/tmp/input
bs = 1024
skip = 7
if = /tmp/input
The code is the same as above, but the two lines are replaced out:
parser.LongPrefix("/");
parser.LongSeparator(":");
% ./test /help
./test {OPTIONS}
This command likes to break your disks
OPTIONS:
/help Show this help menu.
/bs:[BYTES] Block size
/skip:[BYTES] Bytes to skip
/if:[BLOCK SIZE] Block size
/of:[BLOCK SIZE] Block size
% ./test /bs:72 /skip:87 /if:/tmp/test.txt
bs = 72
skip = 87
if = /tmp/test.txt
%
#include <iostream>
#include <args.hxx>
int main(int argc, char **argv)
{
args::ArgumentParser parser("This is a test program.", "This goes after the options.");
args::Group xorgroup(parser, "this group provides xor validation:", args::Group::Validators::Xor);
args::Flag a(xorgroup, "a", "test flag", {'a'});
args::Flag b(xorgroup, "b", "test flag", {'b'});
args::Flag c(xorgroup, "c", "test flag", {'c'});
args::Group nxor(xorgroup, "this group provides all-or-none (nxor) validation:", args::Group::Validators::AllOrNone);
args::Flag d(nxor, "d", "test flag", {'d'});
args::Flag e(nxor, "e", "test flag", {'e'});
args::Flag f(nxor, "f", "test flag", {'f'});
args::Group nxor2(nxor, "this group provides all-or-none (nxor2) validation:", args::Group::Validators::AllOrNone);
args::Flag i(nxor2, "i", "test flag", {'i'});
args::Flag j(nxor2, "j", "test flag", {'j'});
args::Flag k(nxor2, "k", "test flag", {'k'});
args::Group nxor3(nxor, "this group provides all-or-none (nxor3) validation:", args::Group::Validators::AllOrNone);
args::Flag l(nxor3, "l", "test flag", {'l'});
args::Flag m(nxor3, "m", "test flag", {'m'});
args::Flag n(nxor3, "n", "test flag", {'n'});
args::Group atleastone(xorgroup, "this group provides at-least-one validation:", args::Group::Validators::AtLeastOne);
args::Flag g(atleastone, "g", "test flag", {'g'});
args::Flag o(atleastone, "o", "test flag", {'o'});
args::HelpFlag help(parser, "help", "Show this help menu", {'h', "help"});
try
{
parser.ParseCLI(argc, argv);
}
catch (args::Help)
{
std::cout << parser;
return 0;
}
catch (args::ParseError e)
{
std::cerr << e.what() << std::endl;
parser.Help(std::cerr);
return 1;
}
catch (args::ValidationError e)
{
std::cerr << e.what() << std::endl;
parser.Help(std::cerr);
return 1;
}
return 0;
}
% /tmp/test -h
/tmp/test {OPTIONS}
This is a test program.
OPTIONS:
this group provides xor validation:
-a test flag
-b test flag
-c test flag
this group provides all-or-none (nxor) validation:
-d test flag
-e test flag
-f test flag
this group provides all-or-none (nxor2) validation:
-i test flag
-j test flag
-k test flag
this group provides all-or-none (nxor3) validation:
-l test flag
-m test flag
-n test flag
this group provides at-least-one validation:
-g test flag
-o test flag
-h, --help Show this help menu
This goes after the options.
%
I haven't written out a long example for this, but here's the test case you should be able to discern the meaning from:
bool ToLowerReader(const std::string &name, const std::string &value, std::string &destination)
{
destination = value;
std::transform(destination.begin(), destination.end(), destination.begin(), ::tolower);
return true;
}
TEST_CASE("Mapping types work as needed", "[args]")
{
std::unordered_map<std::string, MappingEnum> map{
{"default", MappingEnum::def},
{"foo", MappingEnum::foo},
{"bar", MappingEnum::bar},
{"red", MappingEnum::red},
{"yellow", MappingEnum::yellow},
{"green", MappingEnum::green}};
args::ArgumentParser parser("This is a test program.", "This goes after the options.");
args::MapFlag<std::string, MappingEnum> dmf(parser, "DMF", "Maps string to an enum", {"dmf"}, map);
args::MapFlag<std::string, MappingEnum> mf(parser, "MF", "Maps string to an enum", {"mf"}, map);
args::MapFlag<std::string, MappingEnum, ToLowerReader> cimf(parser, "CIMF", "Maps string to an enum case-insensitively", {"cimf"}, map);
args::MapFlagList<std::string, MappingEnum> mfl(parser, "MFL", "Maps string to an enum list", {"mfl"}, map);
args::MapPositional<std::string, MappingEnum> mp(parser, "MP", "Maps string to an enum", map);
args::MapPositionalList<std::string, MappingEnum> mpl(parser, "MPL", "Maps string to an enum list", map);
parser.ParseArgs(std::vector<std::string>{"--mf=red", "--cimf=YeLLoW", "--mfl=bar", "foo", "--mfl=green", "red", "--mfl", "bar", "default"});
REQUIRE_FALSE(dmf);
REQUIRE(args::get(dmf) == MappingEnum::def);
REQUIRE(mf);
REQUIRE(args::get(mf) == MappingEnum::red);
REQUIRE(cimf);
REQUIRE(args::get(cimf) == MappingEnum::yellow);
REQUIRE(mfl);
REQUIRE((args::get(mfl) == std::vector<MappingEnum>{MappingEnum::bar, MappingEnum::green, MappingEnum::bar}));
REQUIRE(mp);
REQUIRE((args::get(mp) == MappingEnum::foo));
REQUIRE(mpl);
REQUIRE((args::get(mpl) == std::vector<MappingEnum>{MappingEnum::red, MappingEnum::def}));
REQUIRE_THROWS_AS(parser.ParseArgs(std::vector<std::string>{"--mf=YeLLoW"}), args::MapError);
}
This should not really be a question you ask when you are looking for an argument-parsing library, but every test I've done shows args as being about 65% faster than TCLAP and 220% faster than boost::program_options.
The simplest benchmark I threw together is the following one, which parses the
command line -i 7 -c a 2.7 --char b 8.4 -c c 8.8 --char d
with a parser that
parses -i as an int, -c as a list of chars, and the positional parameters as a
list of doubles (the command line was originally much more complex, but TCLAP's
limitations made me trim it down so I could use a common command line across
all libraries. I also have to copy in the arguments list with every run,
because TCLAP permutes its argument list as it runs (and comparison would have
been unfair without comparing all about equally), but that surprisingly didn't
affect much. Also tested is pulling the arguments out, but that was fast
compared to parsing, as would be expected.
% g++ -obench bench.cxx -O2 -std=c++11 -lboost_program_options
% ./bench
args seconds to run: 0.895472
tclap seconds to run: 1.45001
boost::program_options seconds to run: 1.98972
%
#undef NDEBUG
#include <iostream>
#include <chrono>
#include <cassert>
#include "args.hxx"
#include <tclap/CmdLine.h>
#include <boost/program_options.hpp>
namespace po = boost::program_options;
using namespace std::chrono;
inline bool doubleequals(const double a, const double b)
{
static const double delta = 0.0001;
const double diff = a - b;
return diff < delta && diff > -delta;
}
int main()
{
const std::vector<std::string> carguments({"-i", "7", "-c", "a", "2.7", "--char", "b", "8.4", "-c", "c", "8.8", "--char", "d"});
const std::vector<std::string> pcarguments({"progname", "-i", "7", "-c", "a", "2.7", "--char", "b", "8.4", "-c", "c", "8.8", "--char", "d"});
// args
{
high_resolution_clock::time_point start = high_resolution_clock::now();
for (unsigned int x = 0; x < 100000; ++x)
{
std::vector<std::string> arguments(carguments);
args::ArgumentParser parser("This is a test program.", "This goes after the options.");
args::ValueFlag<int> integer(parser, "integer", "The integer flag", {'i', "int"});
args::ValueFlagList<char> characters(parser, "characters", "The character flag", {'c', "char"});
args::PositionalList<double> numbers(parser, "numbers", "The numbers position list");
parser.ParseArgs(arguments);
const int i = args::get(integer);
const std::vector<char> c(args::get(characters));
const std::vector<double> n(args::get(numbers));
assert(i == 7);
assert(c[0] == 'a');
assert(c[1] == 'b');
assert(c[2] == 'c');
assert(c[3] == 'd');
assert(doubleequals(n[0], 2.7));
assert(doubleequals(n[1], 8.4));
assert(doubleequals(n[2], 8.8));
}
high_resolution_clock::duration runtime = high_resolution_clock::now() - start;
std::cout << "args seconds to run: " << duration_cast<duration<double>>(runtime).count() << std::endl;
}
// tclap
{
high_resolution_clock::time_point start = high_resolution_clock::now();
for (unsigned int x = 0; x < 100000; ++x)
{
std::vector<std::string> arguments(pcarguments);
TCLAP::CmdLine cmd("Command description message", ' ', "0.9");
TCLAP::ValueArg<int> integer("i", "int", "The integer flag", false, 0, "integer", cmd);
TCLAP::MultiArg<char> characters("c", "char", "The character flag", false, "characters", cmd);
TCLAP::UnlabeledMultiArg<double> numbers("numbers", "The numbers position list", false, "foo", cmd, false);
cmd.parse(arguments);
const int i = integer.getValue();
const std::vector<char> c(characters.getValue());
const std::vector<double> n(numbers.getValue());
assert(i == 7);
assert(c[0] == 'a');
assert(c[1] == 'b');
assert(c[2] == 'c');
assert(c[3] == 'd');
assert(doubleequals(n[0], 2.7));
assert(doubleequals(n[1], 8.4));
assert(doubleequals(n[2], 8.8));
}
high_resolution_clock::duration runtime = high_resolution_clock::now() - start;
std::cout << "tclap seconds to run: " << duration_cast<duration<double>>(runtime).count() << std::endl;
}
// boost::program_options
{
high_resolution_clock::time_point start = high_resolution_clock::now();
for (unsigned int x = 0; x < 100000; ++x)
{
std::vector<std::string> arguments(carguments);
po::options_description desc("This is a test program.");
desc.add_options()
("int,i", po::value<int>(), "The integer flag")
("char,c", po::value<std::vector<char>>(), "The character flag")
("numbers", po::value<std::vector<double>>(), "The numbers flag");
po::positional_options_description p;
p.add("numbers", -1);
po::variables_map vm;
po::store(po::command_line_parser(carguments).options(desc).positional(p).run(), vm);
const int i = vm["int"].as<int>();
const std::vector<char> c(vm["char"].as<std::vector<char>>());
const std::vector<double> n(vm["numbers"].as<std::vector<double>>());
assert(i == 7);
assert(c[0] == 'a');
assert(c[1] == 'b');
assert(c[2] == 'c');
assert(c[3] == 'd');
assert(doubleequals(n[0], 2.7));
assert(doubleequals(n[1], 8.4));
assert(doubleequals(n[2], 8.8));
}
high_resolution_clock::duration runtime = high_resolution_clock::now() - start;
std::cout << "boost::program_options seconds to run: " << duration_cast<duration<double>>(runtime).count() << std::endl;
}
return 0;
}
So, on top of being more flexible, smaller, and easier to read, it is faster than the most common alternatives.