A self-contained, header-only, collection of some essential C++ utilities that I use extensively for data structure design and benchmarking. Meant to be used in larger projects.

The library is used in several other C++ libraries, such as

• Tongrams, a library for compressed n-gram indexes;
• PTHash, a library implementing efficient minimal and perfect hash functions;
• Auto-Complete, a library for fast and effective query auto-completion;
• Bit-Sliced Indexes, a library providing fast compressed bitmaps.

Integrating Essentials in your own project is very simple: just get the source code and include the header include/essentials.hpp in your code. No other configurations are needed.

If you use git, the easiest way to add Essentials is via git add submodule as follows.

git submodule add https://github.com/jermp/essentials.git

If you are using CMake, you can include the project as follows:

add_subdirectory(path/to/essentials)
target_link_libraries(YourTarget PRIVATE ESSENTIALS)

See also this example repository.

• Benchmarking facilities, including: messages displaying local time, configurable timer class, function to prevent code elision by compiler, simple creation and printing of json documents.

• Functions to serialize-to and load-from disk data structures.

• Functions to compute the number of bytes consumed by data structures.

• Support for creating, removing, and iterate inside directories.

• Experimental: transparent support for contiguous memory allocation.

See the test folder for examples. To compile such examples, just type the following commands from the parent directory.

cd test
mkdir build
cd build
cmake ..
make