Alpha ZealPHP is early-stage and under active development. APIs may change between minor versions until v1.0. Feedback and bug reports welcome on GitHub.
Esc
API Index — Namespaces, Packages, Reports, Indices

MimeResolver
in package

Multi-suffix MIME metadata resolver — Apache mod_mime find_ct parity.

Apache walks EVERY dot-separated suffix of a filename left-to-right (mod_mime.c find_ct, ~874–1007), accumulating Content-Type, Content-Encoding, Content-Language and charset from each suffix. PHP's pathinfo(…, PATHINFO_EXTENSION) only ever returns the rightmost suffix, which is wrong for files such as document.html.gz (HTML body carried with a gzip Content-Encoding) or page.fr.html (French HTML).

This resolver reproduces Apache's algorithm exactly:

  • Basename rule (find_ct lines 874–887): leading dots are part of the basename, and the segment up to the first real dot is the basename and carries no metadata. So .png is a hidden file named png with zero extensions — NO type is assigned (the M12 dotfile fix).
  • Suffix walk (find_ct lines 891–1007): each remaining suffix is lowercased and looked up independently. Empty suffixes ("bad..html") are skipped (line 898).
  • Content-Type: last matching suffix wins — Apache calls ap_set_content_type per match, overwriting (line 921/930).
  • Content-Language: every matching suffix is accumulated in order (lines 938–946, apr_array_push).
  • Content-Encoding: every matching suffix is accumulated in order, comma-joined; duplicates and double-encoding are intentionally preserved (lines 947–962, the -- nd comment).

The resolver is pure: it takes three case-insensitive extension→value maps (type, encoding, language) and a filename, and returns the resolved metadata. It performs no I/O and never inspects file contents.

Table of Contents

Properties

$encodingMap  : array<string, string>
$languageMap  : array<string, string>
$typeMap  : array<string, string>

Methods

__construct()  : mixed
resolve()  : array{type: ?string, encoding: ?string, languages: list}
Walk every suffix of $filename and accumulate metadata.
suffixes()  : array<int, string>
Decompose a filename into its lowercased suffix list, applying Apache's basename rule (leading dots + segment before the first real dot are the basename and excluded). Empty suffixes are dropped.
normaliseMap()  : array<string, string>
Normalise a caller-supplied extension map: lowercase + dot-strip keys, stringify values (Apache lowercases stored values; we preserve the caller's value casing, matching MimeTypeMiddleware's prior behaviour).

Properties

$encodingMap

private array<string, string> $encodingMap

ext => content-encoding (keys lowercased, dot-stripped)

$languageMap

private array<string, string> $languageMap

ext => content-language (keys lowercased, dot-stripped)

$typeMap

private array<string, string> $typeMap

ext => mime-type (keys lowercased, dot-stripped)

Methods

__construct()

public __construct([array<string, string|int> $typeMap = [] ][, array<string, string|int> $encodingMap = [] ][, array<string, string|int> $languageMap = [] ]) : mixed
Parameters
$typeMap : array<string, string|int> = []

ext => mime-type

$encodingMap : array<string, string|int> = []

ext => content-encoding (e.g. gz => gzip)

$languageMap : array<string, string|int> = []

ext => content-language (e.g. fr => fr)

resolve()

Walk every suffix of $filename and accumulate metadata.

public resolve(string $filename) : array{type: ?string, encoding: ?string, languages: list}
Parameters
$filename : string

basename or full path; only the basename's suffix chain is considered.

Return values
array{type: ?string, encoding: ?string, languages: list} —

type is null when no suffix mapped a Content-Type; encoding is the comma-joined encoding chain (null when none); languages is the ordered list of matched languages (empty when none).

suffixes()

Decompose a filename into its lowercased suffix list, applying Apache's basename rule (leading dots + segment before the first real dot are the basename and excluded). Empty suffixes are dropped.

public suffixes(string $filename) : array<int, string>

document.html.gz => ['html', 'gz'] .png => [] (hidden file, no extension) archive.TAR.GZ => ['tar', 'gz'] noext => []

Parameters
$filename : string
Return values
array<int, string>

normaliseMap()

Normalise a caller-supplied extension map: lowercase + dot-strip keys, stringify values (Apache lowercases stored values; we preserve the caller's value casing, matching MimeTypeMiddleware's prior behaviour).

private static normaliseMap(array<string, string|int> $map) : array<string, string>
Parameters
$map : array<string, string|int>
Return values
array<string, string>
On this page