Technically we can write directly from one file to the other. We allocate character buffers in order to make the read/writes faster. But is there a way to maintain the read/write speed while allocating little to no extra memory on the Heap? e.g. like gzip?

Currently you have to build the binaries using terrible instructions in the README.

It contains no member variables, is there a reason to not have this class abstract and avoid unnecessary variable declarations: e.g. BinaryEncoder be() ?

Currently we allocate a ton of memory for the BinaryEncoder object.

It has 4 interface methods:
streamCharacterFileIn (reads in a text files and stores the bits as bools).
streamBinaryFileOut (streams out the currently held bits).
streamBinaryFileIn (streams in bits from a bin file and stores the bits as bools).
streamCharacterFileOut (streams out stored bits into a character file).

Logistically, streamCharacterFileIn and streamBinaryFileOut are called directly with each other in compress.cpp, and the same is true for the latter two methods in decompress.cpp. From an efficiency standpoint, we don't need to actually store any bits in the BinaryEncoder. It could just read from one file, and write to another (using a previously allocated character buffer in between). Storing the bits as bools however would require 8x more memory as bools are stored as bytes anyway!

Find a way to merge .bin and .csm so that we don't need two separate files. Then we can have:
bin/compress file -> file.bin, bin/decompress file.bin -> file

Specifically the description of the algorithm/usage of the binaries. Also include ideas for future development.