Class: RelatonBib::FormattedString

Inherits:
LocalizedString show all
Defined in:
lib/relaton_bib/formatted_string.rb

Overview

Formatted string

Direct Known Subclasses

BiblioNote, FormattedRef

Constant Summary collapse

FORMATS = 
%w[text/plain text/html application/docbook+xml
application/tei+xml text/x-asciidoc text/markdown
application/x-metanorma+xml].freeze

Constants included from RelatonBib

VERSION

Instance Attribute Summary collapse

Attributes inherited from LocalizedString

#content, #language, #script

Instance Method Summary collapse

Methods inherited from LocalizedString

#empty?, #escp, #to_s

Methods included from RelatonBib

array, format_date, grammar_hash, parse_date, parse_yaml

Methods included from Config

#configuration, #configure

Constructor Details

#initialize(content: "", language: nil, script: nil, format: "text/plain") ⇒ FormattedString

Returns a new instance of FormattedString.

Parameters:

  • content (String, Array<RelatonBib::LocalizedString>) (defaults to: "")
  • language (String, nil) (defaults to: nil) —

    language code Iso639

  • script (String, nil) (defaults to: nil) —

    script code Iso15924

  • format (String) (defaults to: "text/plain") —

    the content type



19
20
21
22
23
24
25
26
 
# File 'lib/relaton_bib/formatted_string.rb', line 19

def initialize(content: "", language: nil, script: nil, format: "text/plain")
  # if format && !FORMATS.include?(format)
  #   raise ArgumentError, %{Format "#{format}" is invalid.}
  # end

  @format = format
  super(content, language, script)
end

Instance Attribute Details

#format ⇒ String (readonly)

Returns:

  • (String) 


13
14
15
 
# File 'lib/relaton_bib/formatted_string.rb', line 13

def format
  @format
end

Instance Method Details

#cleanup(str) ⇒ String

Remove HTML tags except <em>, <strong>, <stem>, <sup>, <sub>, <tt>,
, <p>. Replace <i> with <em>, <b> with <strong>.

Parameters:

  • str (String) —

    content

Returns:

  • (String) —

    cleaned content



119
120
121
122
123
124
125
126
127
 
# File 'lib/relaton_bib/formatted_string.rb', line 119

def cleanup(str)
  return str unless format == "text/html"

  str.gsub(/(?<=<)\w+:(?=\w+>)/, "").gsub(/(?<=<\/)\w+:(?=\w+>)/, "")
    .gsub(/<i>/, "<em>").gsub(/<\/i>/, "</em>")
    .gsub(/<b>/, "<strong>").gsub(/<\/b>/, "</strong>")
    .gsub(/<(?!\/?(em|strong|stem|sup|sub|tt|br\s?\/|p))[^\s!]\/?.*?>/, "")
    .gsub(/\s+([.,:;!?<])/, "\\1").strip.squeeze(" ")
end

#encode(cnt) ⇒ String

Encode content.

Parameters:

  • cnt (String) —

    content

Returns:

  • (String) —

    encoded content



41
42
43
44
45
46
47
48
49
50
51
 
# File 'lib/relaton_bib/formatted_string.rb', line 41

def encode(cnt) # rubocop:disable Metrics/MethodLength
  return escp(cnt) unless format == "text/html"

  parts = cnt.scan(%r{
    <(?<tago>\w+)(?<attrs>[^>]*)> | # tag open
    </(?<tagc>\w+)> | # tag close
    (?<cmt><!--.*?-->) | # comment
    (?<cnt>.+?)(?=<|$) # content
    }x)
  scan_xml parts
end

#scan_xml(parts) ⇒ String

Scan XML and escape HTML entities.

Parameters:

Returns:

  • (String) —

    output string



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
 
# File 'lib/relaton_bib/formatted_string.rb', line 60

def scan_xml(parts) # rubocop:disable Metrics/AbcSize,Metrics/CyclomaticComplexity,Metrics/PerceivedComplexity,Metrics/MethodLength
  return "" unless parts.any?

  out = ""
  while parts.any? && (parts.first[3] || parts.first[4])
    _, _, _, cmt, cnt = parts.shift
    out += "#{cmt}#{escp(cnt)}"
  end
  unless out.empty?
    out += scan_xml(parts) if parts.any? && parts.first[0]
    return out
  end

  tago, attrs, tagc, = parts.shift
  out = if tago && attrs && attrs[-1] == "/"
          "<#{tago}#{attrs}>"
        elsif tago
          inr = scan_xml parts
          _, _, tagc, = parts.shift
          if tago == tagc
            "<#{tago}#{attrs}>#{inr}</#{tagc}>"
          else
            "#{escp("<#{tago}#{attrs}>")}#{inr}#{escp("</#{tagc}>")}"
          end
        end
  out += scan_xml(parts) if parts.any? && (parts.first[0] || parts.first[3] || parts.first[4])
  out
end

#to_asciibib(prefix = "", count = 1, has_attrs = false) ⇒ String

Parameters:

  • prefix (String) (defaults to: "")
  • count (Integer) (defaults to: 1) —

    number of elements

Returns:

  • (String) 


102
103
104
105
106
107
108
109
 
# File 'lib/relaton_bib/formatted_string.rb', line 102

def to_asciibib(prefix = "", count = 1, has_attrs = false)
  has_attrs ||= !(format.nil? || format.empty?)
  pref = prefix.empty? ? prefix : "#{prefix}."
  # out = count > 1 ? "#{prefix}::\n" : ""
  out = super
  out += "#{pref}format:: #{format}\n" if format
  out
end

#to_hash ⇒ Hash

Returns:

  • (Hash) 


90
91
92
93
94
95
96
97
 
# File 'lib/relaton_bib/formatted_string.rb', line 90

def to_hash
  hash = super
  return hash unless format

  hash = { "content" => hash } unless hash.is_a? Hash
  hash["format"] = format
  hash
end

#to_xml(builder) ⇒ Object

Parameters:

  • builder (Nokogiri::XML::Builder) 


29
30
31
32
 
# File 'lib/relaton_bib/formatted_string.rb', line 29

def to_xml(builder)
  builder.parent["format"] = format if format
  super
end