kicken/osm-parser

Library for parsing OSM data dump files.

dev-master 2023-09-28 14:45 UTC

Requires

• php: ^8.0

Requires (Dev)

• ext-xmlreader: *
• google/protobuf: ^3.22

Suggests

• ext-xmlreader: Required for XML file parsing
• ext-zlib: Required for PBF files with compressed data.
• google/protobuf: Required for PBF file parsing

MIT

• Keith Maika

• parser
• open street map
• osm
• osm.pbf

README

Parse Open Street Map XML or PBF files with a simple iterator interface.

Example

Extract all named nodes from the Austin, TX dump.

$iter = new Kicken\OSMParser\Parser\XML('Austin.osm');
foreach ($iter as $entity){
    if ($entity instanceof Kicken\OSMParser\Data\Node && $entity->getTag('name')){
        echo "\t", $entity->getTag('name'), ' @ (', $entity->getLat(), ',', $entity->getLon(), ')', PHP_EOL;
    }
}

Resource requirements

While PBF files are much smaller on disk, they require significantly more resources to parse. I tested both parsers using the example code above using PHP 8.2.3 on a Intel i7-1260P laptop and had the following results:

Parser	Time	Peak memory usage
XML	1m 01s	4.1 MB
PBF (Native)	0m 41s	3.1 GB
PBF (PHP)	3m 23s	3.9 GB

XML format is quick and uses the least amount of memory since it can stream as they are found rather than loading a bunch into memory. PBF format is quick if you use the native extension, but requires much more memory as many nodes must be loaded into memory first then output. PBF without the native extension is slow and uses a lot of memory and thus should be avoided.