For many projects it is required that some convention of data storage is used. CSV (comma-separated values) files are an efficient, lightweight, and portable data format that can be used for data storage. Parsing files in C can be tedious, error-prone, and can add unnecessary overhead. This library solves all of these problems by creating a simple struct CSV that can represent the data in a CSV file with ease. Beyond data storage, there are also applications for data analysis, without having to use high-level API's or languages such as Python.
The main header file, csv.h contains all of the necessary structs and function declarations for CSV operations.
csv.c must be included in the project, as it contains the function definitions. Creating binaries for csv.c is a goal for the future.
The utilities header file csvUtils.h offers some common operations that one might wish to do on a CSV file/struct.
Similarly, csvUtils.c contains the function definitions.
To include this library in your project, simply download/clone csv.c and csv.h, and optionally csvUtils.c and csvUtils.h, into your projects and include the header files in your main file.
To learn more about the functions, see the documentation.
An issue that could prevent scaling in the future would be the structure of a CSV file. Having every entry in the CSV taking up the same amount of space means that a file where every entry is one character except for one entry which is 100 characters means the size and efficiency would be drastically reduced. Potential solutions would be:
- Making each entry a pointer to a dynamically allocated string. This would require more referencing, a lot of refactoring, and more init/de-init.
- Making each row have a different Max Entry Size. This would require less refactoring, and would probably minimize the aforementioned issue, but it is not guaranteed nor as effective as the previous option.
- Making the CSV more like an array of strings, where each entry terminates with a null character that signifies the next entry starting. Easy implementation but would greatly reduce performance and add complexity.
- Making the CSV into a linked list. Requires complete refactoring and some worsening of performance but much more efficient memory usage.
- Each entry is its own node and points to the entries right and down of it. 2D (probably singly) linked list.
- Each row (could also be column, it is arbitrary) is a node in a doubly linked list. Each node contains a singly linked list of its entries.
- Each entry is its own node and points to the next entry in a 1D singly linked list.