cpputils

various c++ utility classes

• iostreams based hexdumper utility
• string formatter using variadic templates instead of varargs, making use of iostreams for formatting.
• utf-X string converter.
• c++17 required.
• a argc/argv argument parsing library: argparse.h
• arrayview.h - provide an interface similar to std::stringview, or boost::span
• `asn1parser: decodes BER encoded data
• datapacking: big/little endian data extraction
• fhandle: c++ wrapper for a POSIX file handle.
• fslibrary: enumerates files recursively.
• mmem: memory mapped files.
• stringconvert: utf-N conversion tools.
• stringlibrary: type independent string functions.
• xmlparser: idea based on the python html.parser module.

hexdumper

Using iostream to configure hexdump output:

std::cout << Hex::hexstring << Hex::dumper(data, size) << "\n";
std::cout << Hex::offset(0x12000) << Hex::right << Hex::dumper(data, size) << "\n";

A more detailed description can be found in this blog post

formatter

printf like formatting using a combination of variadic templates and iostreams.

Example:

std::cout << stringformat("%d %s %d", 1LL, std::string("test"), size_t(3));

You can stringify custom types by defining a suitable operator<<(os, customtype).

Compared to alternatives like fmtlib, boost::format, this implementation creates very small binaries. Performance is below that of fmtlib, but well about boost::format.

The code is centered around the StringFormatter class. Several functions use this to provide formatting:

• std::string stringformat(const char *fmt, ...)
  • print to a stl string
• QString qstringformat(const char *fmt, ...)
  • print to a QT string
• fprint(FILE *out, const char*fmt, ...)
  • print to a file
• print(const char*fmt, ...)
  • print to stdout
• debug(const char*fmt, ...)
  • print to windows debug log

benchmarks

Makefile.bench builds several small programs for comparing my formatter to several other similar libraries.

stringconvert

Convert C or C++ strings to C++ strings, between any of UTF-8, UTF-16, UTF-32, depending on the size of the character type.

Example:

auto utf8str = string::convert<char>(L"test");
auto utf16str = string::convert<uint16_t>(L"test");

The first line converts a wchar_t string, which is either utf-16 or utf-32 encoded depending on the compiler, to a utf-8 string, the second line converts to utf-16.

argparse

Class for conveniently parsing commandline arguments.

This example will parse: cmd -apple -a 1234 first second -

for (auto& arg : ArgParser(argc, argv))
   switch (arg.option())
   {
   case 'a': if (arg.match("-apple")) {
                 /*...*/
             }
             else {
                 a_arg = arg.getint();
             }
             break;
   case '-': if (arg.match("--verbose"))
                 verbosity = 1;
             break;
   case 'v': verbosity = arg.count();
             break;
   case 0:   usestdin = true;
             break;
   case -1:  switch(n++)
             {
             case 0: first = arg.getstr(); break;
             case 1: second = arg.getstr(); break;
             }
   }

stringlibrary

Several utility functions for handling NUL terminated char and wchar strings, and std::strings:

• stringcopy, stringlength, stringcompare, stringicompare, stringsplitter, (lr)strip

Parsing integers from strings:

• parseunsigned, parsesigned

datapacking

Classes for packing and unpacking fixed width numeric data, in either little or big-endian format.

fhandle

Exeption safe wrapper for posix filehandles.

mmem

class for using mem-mapped files.

fslibrary

A Recursive file iterator, which can be used from a ranged-for-loop.

asn1parser

Provides several methods of accessing items in an asn.1 BER encoded object.

• asn1tlv is an object which decodes the Type + Length fields, and Provides the 'range' where the Data is located.
• enumtlvs iterates over all BER objects found in the given range.
• traverse extracts a specific part from a BER encoded object.

unittests

A Makefile is provided for building the unittests. These can be build 'normally', and also with options for code coverage testing.

several targets exist:

• all, the old method of building the tests
• cmake, using cmake to resolve dependencies
• vc, for building this on windows platforms with cmake + visualc++

todo

• add support for hexdumping data from streams.
• string alignment / width does not work correctly for unicode characters > 0x80.
• "% d" : space-for-positive is not supported.
• "%\*d" : width from argument list is not supported.
• "%+08d" produces the wrong result: the sign will be after the padding, instead of in front.
• "%.8s" string truncation does not work.
• add support for different alphabets in base32/base64.

(C) 2016 Willem Hengeveld [email protected]