ajgl/csv-rfc

Drop in replacement for native PHP CSV related functions to read and/or write RFC4180 compliant CSV files

AjglCsvRfc

The AjglCsvRfc component offers a drop in replacement for native PHP CSV related functions to read and/or write RFC4180 compliant CSV files., _(*1)

, _(*2)

The native PHP implementation contains a Wont fix bug [#50686] when you try to write a CSV field which contains the escape char (\ by default), followed by the enclosure char (" by default)., _(*3)

The [RFC 4180] states that:, _(*4)

If double-quotes are used to enclose fields, then a double-quote appearing inside a field must be escaped by preceding it with another double quote., _(*5)

The CSV version of the string "Hello\", World! should be """Hello\"", World!" but it does not work as expected. You can see a detailed explanation at https://3v4l.org/NnHp4, _(*6)

This package provides an alternative implementation to read and write well escaped CSV files for the following functions and methods:, _(*7)

Installation

To install the latest stable version of this component, open a console and execute the following command:, _(*8)

$ composer require ajgl/csv-rfc

Usage

Alternative functions

The simplest way to use this library is to call the alternative CSV functions:, _(*9)

use Ajgl\Csv\Rfc;

$handler = fopen('php://temp', 'w+');
Rfc\fputcsv($handler, array('Hello \"World"!'));
rewind($handler);
$row = Rfc\fgetcsv($handler);
rewind($handler);
$row = Rfc\str_getcsv(fgets($handler));

Alternative clases

If you prefer you can use the alternative implementation for SplFileObject or SplTempFileObject:, _(*10)

use Ajgl\Csv\Rfc;

$file = new Rfc\Spl\SplFileObject('php://temp', 'w+');
$file->fputcsv(array('Hello \"World"!'));
$file->rewind();
$row = $file->fgetcsv();
$file->rewind();
$file->setFlags(\SplFileObject::READ_CSV | \SplFileObject::READ_AHEAD | \SplFileObject::SKIP_EMPTY);
foreach ($file as $line) {
    $row = $line;
}

Stream filter

Instead of using the alternative functions or classes, you can use the provided stream filter to fix the enclosure escape. You must register the stream filter (if not registered yet) and append it to your stream:, _(*11)

use Ajgl\Csv\Rfc;

Rfc\CsvRfcWriteStreamFilter::register();

$handler = fopen('php://temp', 'w+');
stream_filter_append(
    $handler,
    Rfc\CsvRfcWriteStreamFilter::FILTERNAME_DEFAULT,
    STREAM_FILTER_WRITE
);
fputcsv($handler, array('Hello \"World"!'));
rewind($handler);
$row = fgetcsv($handler, 0, ',', '"', '"');

❮ NOTE ❯: The $escape_char in [fputcsv] MUST be (if allowed) the default one (the backslash \). The $enclosure and $escape parameters in [fgetcsv] MUST be equals., _(*12)

Custom enclosure character

By default, the enclosure character of the stream filter is a double-quote ("). If you want to change it, you can provide a custom enclosure character in two different ways., _(*13)

Via filter params

An array with an enclosure key can be provided when appending the filter to the stream:, _(*14)

use Ajgl\Csv\Rfc;

$enclosure = '@';
Rfc\CsvRfcWriteStreamFilter::register();

$handler = fopen('php://temp', 'w+');
stream_filter_append(
    $handler,
    Rfc\CsvRfcWriteStreamFilter::FILTERNAME_DEFAULT,
    STREAM_FILTER_WRITE,
    array(
        'enclosure' => $enclosure
    )
);
fputcsv($handler, array('Hello \"World"!'), ',', '@');
rewind($handler);
$row = fgetcsv($handler, 0, ',', '@', '@');

Via filter name

If the filter name starts with the special key csv.rfc.write. you can define your custom enclosure character appending it to the filtername:, _(*15)

use Ajgl\Csv\Rfc;

$enclosure = '@';
$filtername = 'csv.rfc.write.' . $enclosure;
Rfc\CsvRfcWriteStreamFilter::register($filtername);

$handler = fopen('php://temp', 'w+');
stream_filter_append(
    $handler,
    $filtername,
    STREAM_FILTER_WRITE
);
fputcsv($handler, array('Hello \"World"!'), ',', '@');
rewind($handler);
$row = fgetcsv($handler, 0, ',', '@', '@');

❮ NOTE ❯: The enclosure character passed via parameters will override the one defined via filter name., _(*16)

End of line (EOL)

Writing CSV

By default, the PHP CSV implementation uses LF ("\n") as EOL while writing a CSV row. These alternative functions use LF ("\n") by default too., _(*17)

But, the RFC 4180 states that:, _(*18)

Each record is located on a separate line, delimited by a line break (CRLF)., _(*19)

So, if you want to write RFC4180 compliant CSV, you can override the default EOL using:, _(*20)

use Ajgl\Csv\Rfc\CsvRfcUtils;

CsvRfcUtils::setDefaultWriteEol(CsvRfcUtils::EOL_WRITE_RFC);

Reading CSV

To read the CSV data, this implementation leverages the PHP native capabilities to read files. If you are having any problem with the EOL detection, you should enable the auto_detect_line_endings configuration option as following the PHP doc recomendation., _(*21)

ini_set('ini.auto-detect-line-endings', true);

Integration with league/csv <9.0

The well known league/csv package provide a great object oriented API to work with CSV data, but as long as it leverages the default PHP implementation for CSV, versions prior to 9.0 are affected by the [#50686] bug., _(*22)

You can use this component with league/csv <9.0 to produce [RFC 4180] compatible files avoiding this bug., _(*23)

Writer integration

To integrate this component with the league/csv writer implementation, you can use the Stream Filter API., _(*24)

use Ajgl\Csv\Rfc;
use League\Csv\Writer;

CsvRfcWriteStreamFilter::register();
$writer = Writer::createFromPath('/tmp/foobar.csv');
if (!$writer->isActiveStreamFilter()) {
    throw new \Exception("The Stream Filter API is not active.");
}
$writer->appendStreamFilter(CsvRfcWriteStreamFilter::FILTERNAME_DEFAULT);
$writer->insertOne(array('"Hello\", World!'));

❮ NOTE ❯: Do not override the default writer escape character (\)., _(*25)

Known limitations

• The league/csv package does not support the Stream Filter API when the writer instance is created from a SplFileObject.
• The league/csv implementation will always leverage the standard \SplFileObject::fputcsv to write CSV data, so the fix to write [RFC 4180] compatible files from Ajgl\Csv\Rfc\Spl\SplFileObject::fputcsv will be ignored.

Reader Integration

To read back the CSV data, you can leverage the standard implementation, but you MUST set the reader escape and enclosure characters to the same value., _(*26)

use League\Csv\Reader;

$reader = Reader::createFromPath('/tmp/foobar.csv');
$reader->setEscape($reader->getEnclosure());
foreach ($reader as $row) {
    //...
}

License

This component is under the MIT license. See the complete license in the [LICENSE] file., _(*27)

Reporting an issue or a feature request

Issues and feature requests are tracked in the [Github issue tracker]., _(*28)

Author Information

Developed with ♥ by [Antonio J. García Lagar]., _(*29)

If you find this component useful, please add a ★ in the [GitHub repository page] and/or the [Packagist package page]., _(*30)