CSV Reader

This is an incredibly simplistic CSV reader to easily read and fetch data from CSV files.

Building

Standalone

1. Clone this repo: git clone [email protected]:Kiyoshika/csv-reader.git
2. mkdir build && cd build
3. cmake ..
   • Add -DDEBUG_MODE=ON to compile in debug mode
   • Add -DBUILD_TESTS=ON to build tests
4. make to build the static library libcsvreader.a
5. (optional) make test to run tests if you decided to build them

Including in Another CMake Project

1. Add this repo as a submodule to your project
   • E.g., mkdir ext && cd ext
   • git submodule add [email protected]:Kiyoshika/csv-reader.git
2. In your root CMakeLists.txt:
   • add_subdirectory(ext/csv-reader)
   • target_link_libraries([your target] csvreader)
   • target_include_directories([your target] PUBLIC ${CMAKE_SOURCE_DIR}/ext/csv-reader/include)

Where [your target] is your executable or library that needs this as a dependency.

By default, this will compile the library in release mode.

Usage

Namespace

This object lives in the csv namespace, so either use using namespace csv; or preface with csv::CSVReader.

Creating Object

When calling the constructor, you can pass a custom (single-char) delimiter and (single-char) quote character. By default, these are , and " respectively.

CSVReader csv; // by default using ',' and '"'
CSVReader csv2('|', '\''); // use '|' as delim and single quote as quote char

Null Values

You can optionally set a string that defines a NULL value, e.g., NA, N/A, (null), etc.

This library always treats empty columns as null (e.g., empty strings created when two delimiters are directly next to each other.)

Any index out of bounds will automatically return true if checking for null.

csv.set_null_text("NA"); // any text read as "NA" will be marked as null

// check if a column is null at index
if (csv.column_is_null(index)) { ... }

Opening/Closing Files

When you open() a file, you can choose whether or not to skip the header line. By default, this is true. Every open() must have an opposite close(). This is not included in the destructor since nothing is opened upon construction and is thus the programmer's responsibility.

csv.open("path_to_file.csv"); // by default, skip_header = true
csv.open("path_to_file.csv", false); // read header when calling read_line() next

Reading Lines

You read_line() to read the next line in the CSV file which is internally tokenized. This returns true when data has been read and false otherwise (e.g., at the end of the file)

csv.read_line(); // can be used on-demand

// or, in a while loop
while (csv.read_line()) { ... }

Fetching Values

You can then use either get_column<T>(index) to retrieve and cast the column value to your desired types or use get_and_strip_column(index) to read the column as a string and strip the quotes off the end. Note that get_column<T>(...) will throw an exception if the value can not be cast to the given type.

Alternatively, use get_column_or_default<T>(index, default) to attempt to retrieve a value or return a default value if either the value could not be cast properly OR if the value is null. This can simplify things instead of explicitly checking csv.column_is_null(...). Note that while this is convenient, it will silence any underlying issues with the data, so use at your own risk.

All major numeric types are supported as well as std::string.

If colum index is out of range, either 0 or empty string is returned depending on the type given.

If a delimiter is present inbetween quotes, it is ignored and read as part of the string. E.g., "quote,with,delim" if the delimiter is , and quote is ".

For reading strings, get_and_strip_column(index) is the recommended method to prevent unexpected quotes showing up unless you are certain in the quality of your data.

Full Example

Below is a full example showcasing some of the usage.

#include "CSVReader.hpp"

int main()
{
   CSVReader csv;
   csv.set_null_text("NA");
   
   csv.open("path_to_file.csv");
   
   while (csv.read_line())
   {
      auto col1 = csv.get_column_or_default<uint16_t>(0, 0);
      auto col2 = csv.get_column<double>(1);
      auto col3 = csv.get_column<std::string>(2);
      auto col4 = csv.get_and_strip_column(3); // e.g., "value" -> value
   }
   
   csv.close();
   
   return 0;
}

WARNING:

• This library does not safeguard against exceptions if a cast fails. It's the programmer's responsibility to either validate data or handle exceptions.