Module: Unitsdb::Commands::SiTtlParser

Defined in:
lib/unitsdb/commands/si_ttl_parser.rb

Overview

Parser for SI TTL files

Constant Summary collapse

SI_URI_PREFIX = 
"http://si-digital-framework.org/SI/"

Class Method Summary collapse

Class Method Details

.add_symbols_to_entities(entities, graph, si) ⇒ Object

Add symbols to entities



76
77
78
79
80
81
82
 
# File 'lib/unitsdb/commands/si_ttl_parser.rb', line 76

def add_symbols_to_entities(entities, graph, si)
  entities.each do |entity|
    symbol = RDF::Query.new({ RDF::URI(entity[:uri]) => { si.hasSymbol => :value } })
                       .execute(graph).first&.value&.to_s
    entity[:symbol] = symbol if symbol
  end
end

.extract_base_entities(graph, namespace, skos) ⇒ Object

Extract base entities from graph



46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
 
# File 'lib/unitsdb/commands/si_ttl_parser.rb', line 46

def extract_base_entities(graph, namespace, skos)
  entities = []
  processed_uris = {}

  RDF::Query.new({ entity: { skos.prefLabel => :label } })
            .execute(graph).each do |solution|
    entity_uri = solution.entity.to_s
    next unless entity_uri.start_with?(namespace.to_s)
    next if processed_uris[entity_uri]

    processed_uris[entity_uri] = true

    entity_name = entity_uri.split("/").last
    label = RDF::Query.new({ RDF::URI(entity_uri) => { skos.prefLabel => :value } })
                      .execute(graph).first&.value&.to_s
    alt_label = RDF::Query.new({ RDF::URI(entity_uri) => { skos.altLabel => :value } })
                          .execute(graph).first&.value&.to_s

    entities << {
      uri: entity_uri,
      name: entity_name,
      label: label,
      alt_label: alt_label
    }
  end

  entities
end

.extract_entities_from_ttl(entity_type, graph) ⇒ Object

Extract entities from TTL based on entity type



28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
 
# File 'lib/unitsdb/commands/si_ttl_parser.rb', line 28

def extract_entities_from_ttl(entity_type, graph)
  skos = RDF::Vocabulary.new("http://www.w3.org/2004/02/skos/core#")
  si = RDF::Vocabulary.new("http://si-digital-framework.org/SI#")

  namespace_uri = case entity_type
                  when "units" then "http://si-digital-framework.org/SI/units/"
                  when "quantities" then "http://si-digital-framework.org/quantities/"
                  when "prefixes" then "http://si-digital-framework.org/SI/prefixes/"
                  else return []
                  end

  namespace = RDF::Vocabulary.new(namespace_uri)
  entities = extract_base_entities(graph, namespace, skos)
  add_symbols_to_entities(entities, graph, si) if %w[units prefixes].include?(entity_type)
  entities
end

.extract_identifying_suffix(uri) ⇒ Object

Extract suffix from URI for display



85
86
87
88
89
90
91
92
93
94
95
96
97
 
# File 'lib/unitsdb/commands/si_ttl_parser.rb', line 85

def extract_identifying_suffix(uri)
  return "" unless uri

  # For display, we need to format as exactly like the original
  # This helps format the comma-separated multi-units correctly
  if uri.include?("/units/")
    # Return units/name format for units (without duplicating "units/")
    "units/#{uri.split("/").last}"
  else
    # Otherwise strip the prefix
    uri.gsub(SI_URI_PREFIX, "")
  end
end

.parse_ttl_files(dir) ⇒ Object

Parse TTL files and return RDF graph



15
16
17
18
19
20
21
22
23
24
25
 
# File 'lib/unitsdb/commands/si_ttl_parser.rb', line 15

def parse_ttl_files(dir)
  puts "Parsing TTL files in #{dir}..."
  graph = RDF::Graph.new

  Dir.glob(File.join(dir, "*.ttl")).each do |file|
    puts "  Reading #{File.basename(file)}"
    graph.load(file, format: :ttl)
  end

  graph
end