Skip to content


Repository files navigation

The hardf turtle, n-triples, n-quads, TriG and N3 parser for PHP

hardf is a PHP 7.1+ library that lets you handle Linked Data (RDF). It offers:

Both the parser as the serializer have streaming support.

This library is a port of N3.js to PHP

Triple Representation

We use the triple representation in PHP ported from NodeJS N3.js library. Check for more information

On purpose, we focused on performance, and not on developer friendliness. We have thus implemented this triple representation using associative arrays rather than PHP object. Thus, the same that holds for N3.js, is now an array. E.g.:

$triple = [
    'subject' =>   '',
    'predicate' => '',
    'object' =>    '',
    'graph' =>     '', #optional

Encode literals as follows (similar to N3.js)

'"Tom"@en-gb' // lowercase language
'"1"^^' // no angular brackets <>

Library functions

Install this library using composer:

composer require pietercolpaert/hardf


use pietercolpaert\hardf\TriGWriter;

A class that should be instantiated and can write TriG or Turtle

Example use:

$writer = new TriGWriter([
    "prefixes" => [
        "schema" =>"",
        "dct" =>"",
        "geo" =>"",
        "rdf" => "",
        "rdfs"=> ""
    "format" => "n-quads" //Other possible values: n-quads, trig or turtle

echo $writer->end();

All methods

//The method names should speak for themselves:
$writer = new TriGWriter(["prefixes": [ /* ... */]]);
$writer->addTriple($subject, $predicate, $object, $graphl);
$writer->addPrefix($prefix, $iri);
//Creates blank node($predicate and/or $object are optional)
$writer->blank($predicate, $object);
//Creates rdf:list with $elements
$list = $writer->addList($elements);

//Returns the current output it is already able to create and clear the internal memory use (useful for streaming)
$out .= $writer->read();
//Alternatively, you can listen for new chunks through a callback:
$writer->setReadCallback(function ($output) { echo $output });

//Call this at the end. The return value will be the full triple output, or the rest of the output such as closing dots and brackets, unless a callback was set.
$out .= $writer->end();


Next to TriG, the TriGParser class also parses Turtle, N-Triples, N-Quads and the W3C Team Submission N3

All methods

$parser = new TriGParser($options, $tripleCallback, $prefixCallback);
$parser->parse($input, $tripleCallback, $prefixCallback);

Basic examples for small files

Using return values and passing these to a writer:

use pietercolpaert\hardf\TriGParser;
use pietercolpaert\hardf\TriGWriter;
$parser = new TriGParser(["format" => "n-quads"]); //also parser n-triples, n3, turtle and trig. Format is optional
$writer = new TriGWriter();
$triples = $parser->parse("<A> <B> <C> <G> .");
echo $writer->end();

Using callbacks and passing these to a writer:

$parser = new TriGParser();
$writer = new TriGWriter(["format"=>"trig"]);
$parser->parse("<http://A> <https://B> <http://C> <http://G> . <A2> <https://B2> <http://C2> <http://G3> .", function ($e, $triple) use ($writer) {
    if (!isset($e) && isset($triple)) {
        echo $writer->read(); //write out what we have so far
    } else if (!isset($triple))      // flags the end of the file
        echo $writer->end();  //write the end
        echo "Error occured: " . $e;

Example using chunks and keeping prefixes

When you need to parse a large file, you will need to parse only chunks and already process them. You can do that as follows:

$writer = new TriGWriter(["format"=>"n-quads"]);
$tripleCallback = function ($error, $triple) use ($writer) {
    if (isset($error))
        throw $error;
    else if (isset($triple)) {
        echo $writer->read();
    else if (isset($error)) {
        throw $error;
    } else {
        echo $writer->end();
$prefixCallback = function ($prefix, $iri) use (&$writer) {
    $writer->addPrefix($prefix, $iri);
$parser = new TriGParser(["format" => "trig"], $tripleCallback, $prefixCallback);
$parser->end(); //Needs to be called

Parser options

  • format input format (case-insensitive)
    • if not provided or not matching any options below, then any Turtle, TriG, N-Triples or N-Quads input can be parsed (but NOT the N3)
    • turtle - Turtle
    • trig - TriG
    • contains triple, e.g. triple, ntriples, N-Triples - N-Triples
    • contains quad, e.g. quad, nquads, N-Quads - N-Quads
    • contains n3, e.g. n3 - N3
  • blankNodePrefix (defaults to b0_) prefix forced on blank node names, e.g. TriGWriter(["blankNodePrefix" => 'foo']) will parse _:bar as _:foobar.
  • documentIRI sets the base URI used to resolve relative URIs (not applicable if format indicates n-triples or n-quads)
  • lexer allows usage of own lexer class. A lexer must provide following public methods:
    • tokenize(string $input, bool $finalize = true): array<array{'subject': string, 'predicate': string, 'object': string, 'graph': string}>
    • tokenizeChunk(string $input): array<array{'subject': string, 'predicate': string, 'object': string, 'graph': string}>
    • end(): array<array{'subject': string, 'predicate': string, 'object': string, 'graph': string}>
  • explicitQuantifiers - [...]

Empty document base IRI

Some Turtle and N3 documents may use relative-to-the-base-IRI IRI syntax (see here and here), e.g.

<> <someProperty> "some value" .

To properly parse such documents the document base IRI must be known. Otherwise we might end up with empty IRIs (e.g. for the subject in the example above).

Sometimes the base IRI is encoded in the document, e.g.

@base <http://some.base/iri/> .
<> <someProperty> "some value" .

but sometimes it is missing. In such a case the Turtle specification requires us to follow section 5.1.1 of the RFC3986 which says that if the base IRI is not encapsulated in the document, it should be assumed to be the document retrieval URI (e.g. the URL you downloaded the document from or a file path converted to an URL). Unfortunatelly this can not be guessed by the hardf parser and has to be provided by you using the documentIRI parser creation option, e.g.

parser = new TriGParser(["documentIRI" => "http://some.base/iri/"]);

Long story short if you run into the subject/predicate/object on line X can not be parsed without knowing the the document base IRI.(...) error, please initialize the parser with the documentIRI option.


use pietercolpaert\hardf\Util;

A static class with a couple of helpful functions for handling our specific triple representation. It will help you to create and evaluate literals, IRIs, and expand prefixes.

$bool = isIRI($term);
$bool = isLiteral($term);
$bool = isBlank($term);
$bool = isDefaultGraph($term);
$bool = inDefaultGraph($triple);
$value = getLiteralValue($literal);
$literalType = getLiteralType($literal);
$lang = getLiteralLanguage($literal);
$bool = isPrefixedName($term);
$expanded = expandPrefixedName($prefixedName, $prefixes);
$iri = createIRI($iri);
$literalObject = createLiteral($value, $modifier = null);

See the documentation at for more information.

Two executables

We also offer 2 simple tools in bin/ as an example implementation: one validator and one translator. Try for example:

curl -H "accept: application/trig" | php bin/validator.php trig
curl -H "accept: application/trig" | php bin/convert.php trig n-triples


We compared the performance on two turtle files, and parsed it with the EasyRDF library in PHP, the N3.js library for NodeJS and with Hardf. These were the results:

#triples framework time (ms) memory (MB)
1,866 Hardf without opcache 27.6 0.722
1,866 Hardf with opcache 24.5 0.380
1,866 EasyRDF without opcache 5,166.5 2.772
1,866 EasyRDF with opcache 5,176.2 2.421
1,866 ARC2 with opcache 71.9 1.966
1,866 N3.js 24.0
3,896,560 Hardf without opcache 40,017.7 0.722
3,896,560 Hardf with opcache 33,155.3 0.380
3,896,560 N3.js 7,004.0
3,896,560 ARC2 with opcache 203,152.6 3,570.808

License, status and contributions

The hardf library is copyrighted by Ruben Verborgh and Pieter Colpaert and released under the MIT License.

Contributions are welcome, and bug reports or pull requests are always helpful. If you plan to implement a larger feature, it's best to discuss this first by filing an issue.