What is that?

This is math expressions parser. It parses expressions into tree of expressions for further evaluation. Currently it provide float-based calculations but by extending Number class one can substitute float by rational numbers to achieve better precision.

Few assumptions been made:

• Target system has contiguous standard ASCII table (compatible with ISO/IEC 8859-1)
• The input is single-byte characters
• No memory limitations. Building of the evaluation tree is done recursively inside of the brackets which can lead to memory overflow exception at certain nesting level (which is handled though).
• Has reasonable performance avoiding duplicate jobs and extensive memory reallocations
• Target compiler supports C++11 features
• Source code must be built on *nix-family platforms (Ubuntu, MacOSX)
• Source code must be cross-platform and can be easily adopted if needed

What is supported?

Math operations

• + add
• - subtract
• * multiply
• / divide
• ^ power

Priorities (from high to low):

• ^ has highest priority after brackets
• * and / have same medium priority
• + and - have same lowest priority

Unary operators

Numbers can have - or + symbol in front of them to indicate if they are negative or positive. E.g. expression 5*-2 means the same as 5*(-2) and expression 5*+2 means the same as 5*2.

Nevertheless expressions 5*--2 is treated as erroneous as well as 5*++2.

Supported symbols

• Symbols:
• . decimal separator
• (, ) brackets
• spaces (ignored)

Fractional numbers can be written as 1. which means 1.0, or .1 which means 0.1

Supported constants

• pi PI number 3.141592653589793
• e Euler's number

Supported functions

• sin Taking sinus() of argument
• cos Taking cosinus() of argument
• tan Taking tangent() of argument
• atan Taking arctanent() of argument
• asin Taking arcsinus() of argument
• acos Taking arccosinus() of argument
• hypot Computes hypotenuse of a right-angled triangle whose legs are x and y. Takes two arguments X and Y.

Functions can be nested, e.g. sin(cos(pi)) is valid input. Functions may have few arguments delimited by , e.g. hypot(sin(1),0.3) is valid input.

Functions and constants may have same signature.

Operation errors

• Division by zero is supported. Currently function returns 0.0 and prints message.
• Malformed expressions are correctly treated. Currently function returns 0.0 and prints message.

How to build

You have CMake 3.0 or higher installed. Tested with g++ 4.8, g++ 5.5, g++ 6.3, g++ 7.3 and clang (LLVM 9.1.0) compilers.

Checkout repo with git clone and the initialise submodules as it depends on Google Test for unit-tests to be run. Then type git submodule update --recursive --init to checkout Google Test dependency.

Once checked out type the following:

mkdir build && cd build && cmake .. && make -j4 or you can just run build.sh and once it builds type cd build to get into build directory.

Then you can run unit-tests by typing: ./unitTests

and run main app by typing: ./exPar