leeoniya/dump-r

a cleaner, leaner mix of print_r() and var_dump()

2.1 2015-09-04 22:05 UTC

Requires

• php: >=5.4.0

Suggests

• jdorn/sql-formatter: Pretty-print dumped SQL strings

MIT

• Leon Sorokin

• debug
• dump
• pretty
• print
• print_r
• var_dump

README

a cleaner, leaner mix of print_r() and var_dump() (MIT Licensed)

Demo: http://o-0.me/dump_r/

Installing

Composer

https://packagist.org/packages/leeoniya/dump-r

{
	"require": {
		"leeoniya/dump-r": "dev-master"
	}
}

Require

require 'dump_r.php';

Using & Config

Use dump_r() as a drop-in replacement for print_r() and var_dump(). It has some additional arguments that control output. The full signature of the function is:

function dump_r($value, $return = false, $html = true, $depth = 1e3, $expand = 1e3);

• $value is the thing you want to dump
• $return determines whether to return the dump rather than output to screen
• $html controls whether the output is text or html
• $depth sets the recursion limit for the dump
• $expand sets the auto-expanded child node depth

There are also two modifier keys that can be used to control how the node expanding/collapsing works:

1. Holding Shift while toggling will expand/collapse the full depth of the node.
2. Hold Ctrl while toggling will expand/collapse all siblings after that node also. This is useful if you have an array of objects/arrays and want to expand all of them to one level simultaneously by clicking just the first one in the group. It works well for deep, complex objects.
3. Shift and Ctrl can be used together.

Double-clicking binary strings will toggle them between mixed hex/ascii and hex-only representations:

Some types of strings can be pretty-printed and additonal rendering options can be tweaked (shown with defaults):

dump_r\Rend::$xml_pretty	= false;	// pretty-print xml strings
dump_r\Rend::$json_pretty	= false;	// pretty-print json strings
dump_r\Rend::$sql_pretty	= false;	// pretty-print sql strings (requires https://github.com/jdorn/sql-formatter)
dump_r\Rend::$recset_tbls	= true;		// recordset detection & table-style output
dump_r\Rend::$trace_info	= true;		// show file & line where dump_r was invoked
dump_r\Rend::$val_space		= 4;		// number of spaces between key and value columns (affects text output only, not html)

Circular reference (recursion) detection and duplicate output is indicated like this for arrays, objects, closures and resources respectively: [*],{*},(*),<*>.

You can re-style all aspects of the html output using CSS, everything is class-tagged.

Extending

Adding your own classifiers & parsers is extremely easy. Here are instructions and two concrete examples of how the String base type can be subtyped. First for displaying EXIF data of jpeg and tiff image paths and then showing row data from CSV file paths.

This array

$stuff = [
	'imgs/img_1771.jpg',
	'data/people.csv',
];

Which would normally dump like this:

Can be dumped like this with subtyping:

To do this, hook the correct core type and provide a function that classifies and processes the raw value, then modifies and returns an instance of Type. Here are the properties that can be modified/augmented:

1. $type->types - Array of subtype string(s) of your choice. These get appended as CSS classes and are also displayed inline.
2. $type->nodes - Array of expandable subnodes to display. Provide null if no subnodes are needed or to retain any subnodes extracted by the core type.
3. $type->length - A string to be displayed at the end of the line, indicating length of subnodes. You can also abuse this param to display other length-esque information (the EXIF example below uses it to display image dimensions inline). Provide null to retain the default length display for the hooked core type.

use dump_r\Type;

// Example 1: dump EXIF data with image filepath strings

Type::hook('_String', function($raw, Type $type, $path) {
	// match path-esque strings (containing '/' or '\') trailed by an
	// EXIF-capable image extension, then verify this file actually exists
	if (preg_match('#[\/]+.+\.(jpe?g|tiff?)$#', $raw) && is_file($raw)) {
		$nodes = $exif = exif_read_data($raw, 0, true);
		$len = $exif['COMPUTED']['Width'] . 'x' . $exif['COMPUTED']['Height'];

		$type->types	= ['image'];
		$type->nodes	= ['EXIF' => $nodes['EXIF']];
		$type->length	= $len;

		return $type;
	}
});

// Example 2: dump CSV records with csv filepath strings

Type::hook('_String', function($raw, Type $type, $path) {
	if (preg_match('#[\/]+.+\.csv$#', $raw) && is_file($raw)) {

		$type->types	= ['csv'];
		$type->nodes	= csv2array($raw);
		$type->length	= count($type->nodes);

		return $type;
	}
});

function csv2array($file) {
	$csv = [];
	$rows = array_map('str_getcsv', file($file));
	$header = array_shift($rows);
	foreach ($rows as $row)
		$csv[] = array_combine($header, $row);
	return $csv;
}

All core types (see src/dump_r/Node dir) can be hooked by their fully namespaced names. For example, if you wanted to further subtype a JSON object string, you would use

Type::hook('_String\\_JSON\\_Object', function($raw, Type $type, $path) {
	// code here
});

Filtering, Marking & Recursion Control

Using the same Type hooks (introduced above) allows you to modify additional aspects of the renderer and iterator.

Skip specific nodes based on their properties or path in the hierarchy

// prevent anything keyd under 'xxx' from dumping
Type::hook('*', function($raw, Type $type, $path) {
	if (end($path) === 'xxx')
		return false;
});

Stop recursion of specific nodes

// prevent arrays keyed under 'c' from dumping sub-nodes
Type::hook('_Array', function($raw, Type $type, $path) {
	if (end($path) === 'c')
		$type->depth = 1;

	return $type;
});

CSS-tag nodes via classes

// tag nodes keyed under `yyy` with addl CSS classes
Type::hook('*', function($raw, Type $type, $path) {
	if (end($path) === 'yyy') {
		$type->classes[] = 'marked';
	}

	return $type;
});