Class: BestType::PcdmTypeLookup

Inherits:
Object
  • Object
show all
Defined in:
lib/best_type/pcdm_type_lookup.rb

Constant Summary collapse

ARCHIVE =
'Archive'
AUDIO =
'Audio'
DATABASE =
'Database'
DATASET =
'Dataset'
EMAIL =
'Email'
FONT =
'Font'
HTML =
'HTML'
IMAGE =
'Image'
PAGE_DESCRIPTION =
'PageDescription'
PRESENTATION =
'Presentation'
SOFTWARE =
'Software'
SOURCE_CODE =
'SourceCode'
SPREADSHEET =
'Spreadsheet'
STRUCTURED_TEXT =
'StructuredText'
TEXT =
'Text'
UNKNOWN =
'Unknown'
UNSTRUCTURED_TEXT =
'UnstructuredText'
VIDEO =
'Video'
WEBSITE =
'Website'
VALID_TYPES =

these include values that will not be derived from MIME/content types

[
  ARCHIVE, AUDIO, DATABASE, DATASET, EMAIL, FONT, HTML, IMAGE, PAGE_DESCRIPTION, PRESENTATION, SOFTWARE,
  SOURCE_CODE, SPREADSHEET, STRUCTURED_TEXT, TEXT, UNKNOWN, UNSTRUCTURED_TEXT, VIDEO, WEBSITE
].freeze
FALLBACK_TYPE =
UNKNOWN

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(mime_type_lookup_instance) ⇒ PcdmTypeLookup

Returns a new instance of PcdmTypeLookup.



40
41
42
43
# File 'lib/best_type/pcdm_type_lookup.rb', line 40

def initialize(mime_type_lookup_instance)
  @mime_type_lookup = mime_type_lookup_instance
  @config = @mime_type_lookup.config
end

Instance Attribute Details

#configObject (readonly)

Returns the value of attribute config.



9
10
11
# File 'lib/best_type/pcdm_type_lookup.rb', line 9

def config
  @config
end

Instance Method Details

#fallback_typeObject



45
46
47
# File 'lib/best_type/pcdm_type_lookup.rb', line 45

def fallback_type
  FALLBACK_TYPE
end

#for_file_name(file_name_or_path) ⇒ Object



53
54
55
56
57
58
# File 'lib/best_type/pcdm_type_lookup.rb', line 53

def for_file_name(file_name_or_path)
  # Normalize format of file_name_or_path
  file_name_or_path = file_name_or_path.downcase

  for_mime_type(@mime_type_lookup.for_file_name(file_name_or_path))
end

#for_mime_type(mime_type) ⇒ Object



60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
# File 'lib/best_type/pcdm_type_lookup.rb', line 60

def for_mime_type(mime_type)
  # Normalize format of mime_type
  mime_type = mime_type.downcase

  # Check config overrides first
  file_type = @config.mime_type_to_pcdm_type_overrides.fetch(mime_type, nil)
  return file_type unless file_type.nil?

  mimes_to_type = {
    /^image/i => IMAGE,
    /^video/i => VIDEO,
    /^audio/i => AUDIO,
    /^text/i => {
      %r{/css}i => SOURCE_CODE,
      %r{/html}i => HTML,
      /.+/ => TEXT
    },
    /excel|spreadsheet|xls/i => SPREADSHEET,
    %r{application/sql}i => DATABASE,
    /csv/i => DATASET,
    /octet.stream/i => UNKNOWN,
    /^application/i => {
      %r{/access}i => DATABASE,
      %r{/css}i => SOURCE_CODE,
      %r{/html}i => HTML,
      %r{/x-iwork-keynote}i => PRESENTATION,
      %r{/x-iwork-numbers}i => SPREADSHEET,
      %r{/x-iwork-pages}i => PAGE_DESCRIPTION,
      %r{/mbox}i => EMAIL,
      %r{/mp4}i => VIDEO,
      %r{/mp4a}i => AUDIO,
      %r{/msaccess}i => DATABASE,
      %r{/mxf}i => VIDEO,
      %r{/(pdf|msword)}i => PAGE_DESCRIPTION,
      %r{/postscript}i => PAGE_DESCRIPTION,
      %r{/powerpoint}i => PRESENTATION,
      %r{/rtf}i => PAGE_DESCRIPTION,
      %r{/sql}i => DATABASE,
      %r{/swf} => VIDEO,
      %r{/vnd.ms-asf}i => VIDEO,
      %r{/vnd.ms-word}i => PAGE_DESCRIPTION,
      %r{/vnd.ms-wpl}i => PAGE_DESCRIPTION,
      %r{/vnd.oasis.opendocument.text}i => PAGE_DESCRIPTION,
      %r{/vnd.openxmlformats-officedocument.presentation}i => PRESENTATION,
      %r{/vnd.openxmlformats-officedocument.wordprocessingml}i => PAGE_DESCRIPTION,
      %r{/vnd.ms-powerpoint}i => PRESENTATION,
      %r{/vnd.sun.xml.calc}i => SPREADSHEET,
      %r{/vnd.sun.xml.impress}i => PRESENTATION,
      %r{/vnd.sun.xml.writer}i => PAGE_DESCRIPTION,
      %r{/xml}i => STRUCTURED_TEXT,
      %r{/x.mspublisher}i => PAGE_DESCRIPTION,
      %r{/x.shockwave-flash} => VIDEO,
      %r{/x.spss}i => DATASET,
      %r{/zip}i => ARCHIVE,
      /.+/ => UNKNOWN
    }
  }

  file_type = mimes_to_type.detect { |pattern, _type_val| mime_type =~ pattern }
  return fallback_type unless file_type

  file_type = file_type.last.detect { |pattern, _type_val| mime_type =~ pattern } if file_type&.last.is_a? Hash
  file_type.nil? ? fallback_type : file_type.last
end

#valid_type?(value) ⇒ Boolean

Returns:

  • (Boolean)


49
50
51
# File 'lib/best_type/pcdm_type_lookup.rb', line 49

def valid_type?(value)
  VALID_TYPES.include? value
end