This is the documentation for the unsupported version 7.0. Please consider upgrading your code to the latest stable version

Managing the BOM character

Detecting the CSV BOM character

To improve interoperability with programs interacting with CSV, you can now manage the presence of a BOM character in your CSV content. The character signals the endianness of the CSV and its value depends on the CSV encoding character. To help you work with BOM, we are adding the following constants to the Reader and the Writer class:

  • BOM_UTF8 : UTF-8 BOM;
  • BOM_UTF16_BE : UTF-16 BOM with Big-Endian;
  • BOM_UTF16_LE : UTF-16 BOM with Little-Endian;
  • BOM_UTF32_BE : UTF-32 BOM with Big-Endian;
  • BOM_UTF32_LE : UTF-32 BOM with Little-Endian;

They each represent the BOM character for each encoding character.


This method will detect and return the BOM character used in your CSV if any.

$reader = new Reader::createFromPath('/path/to/your/file.csv', 'r');
$res = $reader->getInputBOM(); //$res equals null if no BOM is found

$reader = new Reader::createFromPat('path/to/your/msexcel.csv');
if (Reader::BOM_UTF16_LE == $reader->getInputBOM()) {
    //the CSV file is encoded using UTF-16 LE

If you wish to remove the BOM character while processing your data, you can rely on the query filters to do so.

Adding the BOM character to your CSV

setOutputBOM($bom = null)

This method will manage the addition of a BOM character in front of your outputted CSV when you are:

  • downloading a file using the output method
  • outputting the CSV directly using the __toString() method

$bom is a string representing the BOM character. To remove the BOM character just set $bom to an empty value like null or an empty string.

To ease writing the sequence you should use the BOM_* constants.


This method will tell you at any given time what BOM character will be prepended to the CSV content.

For Backward compatibility by default getOutputBOM returns null.

$reader = new Reader::createFromPath('/path/to/your/file.csv', 'r');
$reader->getOutputBOM(); //$res equals null;
$res = $reader->getOutputBOM(); //$res equals "\xFF\xFE";
echo $reader; //the BOM sequence is prepended to the CSV

Software dependency

Depending on your operating system and on the software you are using to read/import your CSV you may need to adjust the encoding character and add its corresponding BOM character to your CSV.

Out of the box, League\Csv assumes that your are using a UTF-8 encoded CSV without any BOM character.

In the examples below we will be using an existing CSV as a starting point. The code may vary if you are creating the CSV from scratch.

MS Excel on Windows

On Windows, MS Excel, expects an UTF-8 encoded CSV with its corresponding BOM character. To fulfill this requirement, you simply need to add the UTF-8 BOM character if needed as explained below:

use League\Csv\Reader;

require '../vendor/autoload.php';

$reader = Reader::createFromPath('/path/to/my/file.csv', 'r');
//BOM detected and adjusted for the output
echo $reader->__toString();

MS Excel on MacOS

On a MacOS system, MS Excel requires a CSV encoded in UTF-16 LE using the tab character as delimiter. Here’s an example on how to meet those requirements using the League\Csv package.

use League\Csv\Reader;
use League\Csv\Writer;
use lib\FilterTranscode;

require '../vendor/autoload.php';

//the current CSV is UTF-8 encoded with a ";" delimiter
$origin = Reader::createFromPath(__DIR__.'/data/prenoms.csv');

//let's convert the CSV to use a tab delimiter.

//we must use a real temp file to be able to rewind the cursor file
//without loosing the modifications
$writer = Writer::createFromPath('/tmp/toto.csv', 'w');

//we set the tab as the delimiter character

//we insert csv data

//let's switch to the Reader object
//Writer::output will failed because of the open mode
$csv = $writer->newReader();

//we register a Stream Filter class to convert the CSV into the UTF-16 LE
stream_filter_register(FilterTranscode::FILTER_NAME."*", "\lib\FilterTranscode");

//we detect and adjust the output BOM to be used
//all is good let's output the results

Of note, we used the filtering capability of the library to convert the CSV encoding character from UTF-8 to UTF-16 LE.

You can found the code and the associated filter class in the examples directory.