`Web::MIME::Sniffer`

MIME type sniffer

SYNOPSIS

  use Web::MIME::Sniffer;
  use Web::MIME::Type;
  my $ct = $response->get_headers_by_name ('Content-Type')->[-1];
  my $mime = defined $ct ? Web::MIME::Type->parse_web_mime_type ($ct) : undef;
  my $sniffer = Web::MIME::Sniffer->new_from_context ('navigate');
  $sniffer->supported_image_types->{'image/jpeg'} = 1;
  $sniffer->supported_image_types->{'image/png'} = 1;
  $sniffer->supported_image_types->{'image/gif'} = 1;
  my $computed_mime = $sniffer->detect (
    $mime,
    (substr $response->body, 0, 1445),
  );
  warn $computed_mime->mime_type_portion;

DESCRIPTION

The Web::MIME::Sniffer class is a MIME type sniffer for Web user agents. It implements the MIME sniffing algorithm of the MIME Sniffing Standard.

METHODS

There are following methods:

$sniffer = Web::MIME::Sniffer->new_from_context ($context)

Create a new instance of the sniffer. The argument must be one of followings:

  navigate       For a normal navigation.
  image          For an image (e.g. fetching <img src>).
  audio_or_video For an audio or video (e.g. fetching <video src>).
  font           For a font (e.g. fetching Web Fonts data).
  text_track     For a text track (e.g. fetching <track src>).
  object         For fetching <object data>.

$sniffer->supported_image_types->{$mime_type_portion} = $boolean

Whether the application supports the MIME type as an image type or not. This can affect the sniffing algorithm's result.

The hash key must be a MIME type portion (type followed by a / followed by a subtype, without parameters) in lowercase, such as image/png or image/jpeg.

This hash reference is empty by default.

$sniffer->supported_audio_or_video_types->{$mime_type_portion} = $boolean

Whether the application supports the MIME type as an audio or video type or not. This can affect the sniffing algorithm's result.

The hash key must be a MIME type portion (type followed by a / followed by a subtype, without parameters) in lowercase, such as image/png or image/jpeg.

This hash reference is empty by default.

$boolean = $sniffer->is_http

$sniffer->is_http ($boolean)

Get or set whether the response in question has been received via HTTP or not. If the underlying protocol is HTTP or FTP over HTTP, this value should be set to true. Otherwise, e.g. if the response is created from a data: URL or from file system, this value should be set to false. Default is false.

$mime = $sniffer->detect ($mime, $bytes)

Run the sniffing alorithm.

The first argument must be a MIME type object (Web::MIME::Type) or undef. If the resource in question is an HTTP response, the MIME type object created from the last Content-Type header value (see the MIME Sniffing Standard) should be specified. If parsing of the header value returns undef, or there is no Content-Type header, undef should be specified.

The second arugment must be the resource header, i.e. first 1445 bytes of the response body (see the MIME Sniffing Standard for details).

The method returns a MIME type object (Web::MIME::Type) representing the computed MIME type.

Please note that when the computed MIME type is different from the supplied MIME type (i.e. the return value is different from the first argument), the computed MIME type does not have parameters included in the supplied MIME type. Parameters, especially the charset parameter, should be taken from the supplied MIME type, as defined by the relevant specifications.

SPECIFICATIONS

MIME Sniffing Standard <https://mimesniff.spec.whatwg.org/>.

HTML Standard <https://html.spec.whatwg.org/>.

<track> Don't check Content-Type for <track> <https://www.w3.org/Bugs/Public/show_bug.cgi?id=13995>.

HISTORY

This module derived from the Whatpm::ContentType module in the manakai-core package <https://github.com/wakaba/manakai>.

AUTHOR

Wakaba <wakaba@suikawiki.org>.

LICENSE

This library is free software; you can redistribute it and/or modify it under the same terms as Perl itself.

The manakai project

Web::MIME::Sniffer