UgAr-study/csv_reader

Build

In project directory:

$ mkdir build
$ cmake build
$ cmake --build build --target csv_manager

Run

In project directory:

$ ./build/csv_manager <path to csv file>

Testing

In directory test there is a python script for running all the .csv tests in this directory. You can run it from test directory:

$ python3 tester.py

Program description

The main class of this program is CSV:

class CSV
{
public:
    CSV() {}
    CSV(const std::string& filename);
    void load_from_file(const std::string& filename);
    void process();
    void save(const std::string& filename);
    void save(std::ostream& os);
};

• CSV::CSV(); CSV::CSV(filename) - constructs an empty CSV or filled with file content.
• CSV::load_from_file(filename) - fills an empty CSV with content from file.
• CSV::process() - calculates all the CSV cells and make it valid for saving.
• CSV::save(filename); CSV::save(ostream) - writes CSV content to the file or to the ostream directly. NOTE: works only after successfull CSV::process.

CSV can be constructed from file directly, or with help of CSV::load_from_file method. After loading from file CSV stores each cell as Expression to make it possible to calculate the cell's value from expression. This calculation is made in CSV::process method, which firstly checks the possibility of making such calculation and only after that computes all the expressions. It is important and convenient for further functionality extension, because the expressions can be modified before invoking the CSV::process.

To extent the functionality one can add new types in the enums: Node_t or BinOp_t, and describe its handling in the following functions: lexer, BuildTree and TreeCalculation.

For example if someone wants to introduce functions, such as MIN, MAX, etc., he can add new type FUNC in Node_t and create new enum Func_t for built-in functions. Then he should write the handling process of new type in lexer, BuildTree and TreeCalculation.