Uni-CSV

Overview

Uni-CSV is a lightweight and efficient CSV parsing utility designed specifically for Unity projects. It allows easy and flexible parsing of CSV data from both files and strings, supporting multiple delimiter types (comma, tab, semicolon, pipe) and handling common edge cases like quoted fields and multiline data.

Supported Unity versions

This package runs on Unity 2019.2 or later.

Installation

This package can be installed with the built-in Unity Package Manager. Open the UPM window, click on the + icon, choose Add package from git URL... and paste this URL:

https://github.com/Sov3rain/Uni-CSV.git?path=/Assets/uni-csv

You can alternatively install a specific version:

https://github.com/Sov3rain/Uni-CSV.git?path=/Assets/uni-csv#1.2.0

Usage

Basic

Returns CSV data as List<List<string>>.

You can parse a string using:

CsvParser.ParseFromString(
    string data, 
    bool hasHeader,
    bool removeHeader = true,
    Delimiter delimiter = Delimiter.Auto)

or a file using:

CsvParser.ParseFromPath(
    string path,
    bool hasHeader,
    bool removeHeader = true,
    Delimiter delimiter = Delimiter.Auto,
    Encoding encoding = null)

Both methods have the header parameter set to true by default. If your CSV file does not contains a header row, set this parameter to false.

Advanced

You can map your CSV to a concrete type using generic methods, which will return an IEnumerator<T>. Keep in mind that for the mapping to work properly, the input string or file must include a header row when using these generic methods.

CsvParser.ParseFromString<T>(
    string data,
    bool hasHeader,
    Delimiter delimiter = Delimiter.Auto)

CsvParser.ParseFromPath<T>(
    string path,
    bool hasHeader,
    Delimiter delimiter = Delimiter.Auto,
    Encoding = null)

Mapping the CSV to a collection of concrete types is performed using reflection, which can affect performance, even though it is only used once on the header row.

Examples

Getting back a List<LIst<string>>:

var sheet = CsvParser.ParseFromString(csvString, hasHeader: true);

foreach (var row in sheet)
{
    Debug.Log(string.Join(", ", row));
}

Getting back a mapped collection of objects:

class User 
{
    public string Username { get; set; }
    public string Email { get; set; }
}

var users = CsvParser.ParseFromString<User>(csvString, hasHeader: true);

foreach (User user in users)
{
    Debug.Log(user.Username);    
}

If the names in the header row don’t match your property names or contain reserved or invalid characters, you can manually map the property to the correct header using the [CsvColumn(name)] and [CsvColumnIndex(index)] attributes:

class User
{
    [CsvColumn("User name")]
    public string Username { get; set; }

    [CsvColumnIndex(1)]
    public string Email { get; set; }
}

var users = CsvParser.ParseFromString<User>(csvString, hasHeader: true);

Both attributes can be mixed in a class, but name-based mapping will be prioritized.

Specs

Compliant with RFC 4180.

• Delimiter auto detection (,, ;, \t and | supported).
• Correctly parse new lines, commas and quotation marks inside cell.
• Escape double quotes.
• Support for some encoding types (default is UTF-8).
• Correctly escapes empty lines and empty content lines.

Roadmap

• Streaming loading
• CSV Writing

Tests

This package has a set of tests that can be run with the Unity Test Framework.

Acknowledgment

This package is freely inspired by GitHub - yutokun/CSV-Parser: CSV Parser for C# without any dependency (on recent platforms).