relaton-render 0.4.2 → 3.0.0.pre.alpha.26

data/lib/relaton/render/iso690/styles/apa.yml

@@ -0,0 +1,45 @@
+# American Psychological Association (7th ed.) reference forms: all
+# author names invert with initials ("Author, A. B."), the date
+# follows the authors in parentheses, article titles take no marks
+# (sentence case is the author's), and only book titles and journal
+# names carry italics. The reference list joins with a comma and ampersand.
+name: "APA (7th edition) references"
+family: apa
+scheme:
+  system: name-date
+  nameForm:
+    initials: true
+    initialsSeparator: " "
+    initialsPeriod: true
+    invertedAll: true
+    givenNameFirst: false
+    editorsMarked: false
+    surnameUpcase: false
+    subsequentSurnameUpcase: false
+  locale:
+    "in": "In"
+    labels:
+      "and": "&"
+      oxford_comma: ","
+      oxford_comma: ","
+templates:
+  titleOpen: ""
+  titleClose: ""
+  citation: "{{surname}}, {{date}}"
+  reference: "{{creator}} ({{date}}). {{title}}. {{series}}. {{production}}. {{extent}}"
+  name: "{{surname}}, {{givenNames}}"
+  identifierMode: all
+perType:
+  - type: report
+    home: true
+    template: "{{title}}"
+  - type: report
+    template: "{{identifier}}. {{title}}. {{production}}."
+  - type: monograph
+    title: "<em>{{title}}</em>"
+    template: "{{creator}} ({{date}}). {{title}}. {{production}}."
+  - type: component_part
+    template: "{{creator}} ({{date}}). {{title}}. {{componentpart}} ({{extent}}). {{production}}."
+  - type: continuing
+    series: "<em>{{seriestitle}}, {{serienumber}}</em>"
+    template: "{{creator}} ({{date}}). {{title}}. {{series}} {{extent}}. {{location}}"

data/lib/relaton/render/iso690/styles/author-date.yml

@@ -0,0 +1,31 @@
+# ISO 690 name-and-date citation style, an instance of the relaton-models
+# CitationStyle model. Templates place the ISO 690 clause 7 data elements
+# ({{creator}}, {{title}}, ...) in each resource kind's clause 8 order; an
+# absent element takes its trailing punctuation with it. A new style is a
+# YAML instance — the engine holds no styles.
+name: "ISO 690 (name and date)"
+scheme:
+  system: name-date
+  nameForm:
+    initials: false
+    givenNameFirst: false
+    # ISO 690 marks an editor list's role ("(éd.)") also when the list
+    # holds a single editor; flavors opt out per their own style
+    loneEditorMarked: true
+templates:
+  titleOpen: "_"
+  titleClose: "_"
+  citation: "{{surname}}, {{date}}{{disambiguator}}"
+  reference: "{{creator}}. {{title}}. {{medium}}. {{edition}}. {{series}}. {{production}}, {{date}}. {{identifier}}. {{location}}."
+  name: "{{surname}}, {{givenNames}}"
+perType:
+  - type: monograph
+    template: "{{creator}}. {{title}}. {{edition}}. {{series}}. {{production}}, {{date}}. {{identifier}}. {{location}}."
+  - type: report
+    template: "{{creator}}. {{title}}. {{numeration}}. {{edition}}. {{series}}. {{production}}, {{date}}. {{identifier}}. {{location}}."
+  - type: patent
+    template: "{{creator}}. {{title}}. {{identifier}}. {{date}}."
+  - type: continuing
+    template: "{{creator}}. {{title}}. {{componentPart}}. {{medium}}. {{edition}}. {{production}}, {{date}}. {{numeration}}. {{identifier}}. {{location}}."
+  - type: online
+    template: "{{creator}}. {{title}}. {{date}}. {{identifier}}. {{location}}."

data/lib/relaton/render/iso690/styles/chicago.yml

@@ -0,0 +1,48 @@
+# Chicago Manual of Style (17th ed.), bibliography forms: the notes
+# system's reference list. Only the first author name inverts ("Peck,
+# R. B., W. E. Hanson, and T. H. Thornburn"), article titles carry
+# quotation marks, book titles are set in italics, and production runs
+# "Place: Publisher".
+name: "Chicago Manual of Style (bibliography)"
+family: chicago
+requires:
+  - extent_units
+scheme:
+  system: name-date
+  nameForm:
+    initials: false
+    initialsSeparator: " "
+    initialsPeriod: true
+    invertedAll: false
+    givenNameFirst: false
+    editorsMarked: true
+    loneEditorMarked: true
+    surnameUpcase: false
+    subsequentSurnameUpcase: false
+  locale:
+    "in": "in"
+    labels:
+      "and": "and"
+      oxford_comma: ","
+      series_no: "no."
+templates:
+  titleOpen: "\""
+  titleClose: "\""
+  citation: "{{surname}}, {{date}}"
+  reference: "{{creator}}. {{title}}. {{edition}}. {{series}}. {{production}}, {{date}}. {{extent}}."
+  name: "{{surname}}, {{givenNames}}"
+  identifierMode: all
+perType:
+  - type: report
+    home: true
+    template: "{{title}}"
+  - type: report
+    template: "{{creator}}. {{title}}. {{edition}}. {{production}}, {{date}}."
+  - type: monograph
+    title: "<em>{{title}}</em>"
+    template: "{{creator}}. {{title}}. {{edition}}. {{production}}, {{date}}. {{extent}}."
+  - type: component_part
+    template: "{{creator}}. \"{{title}}.\" {{componentpart}}. {{production}}, {{date}}. {{extent}}."
+  - type: continuing
+    series: "<em>{{seriestitle}}</em>"
+    template: "{{creator}}. \"{{title}}.\" {{series}} {{extent}} ({{date}}). {{location}}"

data/lib/relaton/render/iso690/styles/ieee-sa.yml

@@ -0,0 +1,45 @@
+# IEEE-SA Standards Style Manual bibliography forms (clause 17): a
+# standard's entry is its designation and title, nothing else - no
+# authors, no publisher, no place; articles and books follow the
+# Chicago-derived forms, with only the first author name inverting and
+# the in-text citation carrying the designation and the [B#] number.
+name: "IEEE-SA standards bibliography"
+family: ieee-sa
+scheme:
+  system: designation
+  nameForm:
+    initials: true
+    initialsSeparator: " "
+    initialsPeriod: true
+    invertedAll: false
+    givenNameFirst: false
+    editorsMarked: false
+    surnameUpcase: false
+    subsequentSurnameUpcase: false
+  locale:
+    labels:
+      "and": "and"
+      oxford_comma: ","
+templates:
+  titleOpen: ""
+  titleClose: ""
+  citation: "{{identifier}}"
+  reference: "{{creator}}, {{title}}. {{series}}. {{extent}}. {{date}}."
+  name: "{{surname}}, {{givenNames}}"
+  identifierMode: first
+perType:
+  - type: report
+    home: true
+    template: "{{title}}"
+  - type: report
+    template: "{{identifier}}, {{title}}"
+  - type: monograph
+    title: "<em>{{title}}</em>"
+    template: "{{creator}}, {{title}}, {{edition}}. {{production}}, {{date}}, {{extent}}."
+  - type: component_part
+    title: "\"{{title}},\""
+    template: "{{creator}}, {{title}} {{componentpart}}, {{extent}}, {{date}}."
+  - type: continuing
+    title: "\"{{title}},\""
+    series: "<em>{{seriestitle}}</em>"
+    template: "{{creator}}, {{title}} {{series}}, {{extent}}, {{date}}."

data/lib/relaton/render/iso690/styles/lncs.yml

@@ -0,0 +1,54 @@
+# Springer LNCS citation style, an instance of the relaton-models
+# CitationStyle model. The LNCS shape: "Author, A.B.: Title. In:
+# Editor, E. (ed.), Proceedings. LNCS 1234, pp. 10-20. Springer,
+# Heidelberg (2010)" - the creator list closes on a colon, the year
+# closes the entry in parentheses, publishers precede places, and a
+# standard with no personal authors leads with the publisher's
+# abbreviation.
+name: "Springer LNCS"
+family: lncs
+requires:
+  - production_order
+  - extent_units
+scheme:
+  system: name-date
+  nameForm:
+    initials: true
+    initialsSeparator: ""
+    initialsPeriod: true
+    invertedAll: true
+    givenNameFirst: false
+    editorsMarked: true
+    loneEditorMarked: true
+    creatorFallback: publisher_abbrev
+    surnameUpcase: false
+    subsequentSurnameUpcase: false
+  locale:
+    "in": "In"
+    availableAt: ""
+    labels:
+      "and": ""
+      oxford_comma: ""
+      series_no: ""
+    punct:
+      production_order: publisher_first
+      production_sep: ", "
+templates:
+  titleOpen: ""
+  titleClose: ""
+  citation: "{{surname}}, {{date}}"
+  reference: "{{creator}}: {{title}}. {{series}}. {{production}} {{dategroup}}. {{location}}"
+  name: "{{surname}}, {{givenNames}}"
+  identifierMode: first
+perType:
+  - type: report
+    home: true
+    template: "{{title}}"
+  - type: report
+    template: "{{creator}}: {{identifier}}, {{title}}. {{production}} {{dategroup}}."
+  - type: monograph
+    template: "{{creator}}: {{title}}. {{series}} {{production}} {{dategroup}}"
+  - type: component_part
+    template: "{{creator}}: {{title}}. {{componentpart}} {{extent}} {{production}} {{dategroup}}"
+  - type: continuing
+    template: "{{creator}}: {{title}}. {{series}} {{extent}} {{dategroup}}"

data/lib/relaton/render/iso690/template.rb

@@ -0,0 +1,248 @@
+# frozen_string_literal: true
+
+module Relaton
+  module Render
+    module Iso690
+      # A citation template as declared in a style instance: literal text
+      # interleaved with {{slot}} placeholders resolved against the ISO 690
+      # data element inventory. A leading literal attaches to the first
+      # slot, and a literal between two slots to the preceding one, so an
+      # absent field never orphans punctuation around it. The last rendered
+      # slot gives up its trailing literal in favour of the template
+      # terminator, appended once when any field rendered and the output
+      # does not already end with it.
+      class Template
+        SLOT = /\{\{\s*([\w-]+)\s*\}\}/.freeze
+
+        Field = Struct.new(:present?, :text)
+
+        def initialize(source)
+          @chunks = []
+          @terminator = ""
+          scan(source.to_s)
+        end
+
+        # The short cite renders the reference with the first-biblio
+        # marker appended to its first component: isodoc's styled
+        # references (eref2linkshort) split there (the 1.x
+        # fmt-first-biblio-delim)
+        def evaluate_short(fields, delim)
+          present = present_indices(fields)
+          return "" if present.empty?
+
+          # The split marker follows the leading run of present slots
+          # (the 1.x first SEGMENT: "{{creator}}{{role}}" splits after
+          # the role), before the first literal that follows it
+          lead_slots = []
+          @chunks.each_with_index do |(kind, _value), i|
+            break if kind != :slot || !present.include?(i)
+
+            lead_slots << i
+          end
+          split_at = lead_slots.last ? lead_slots.last + 1 : nil
+
+          out = +""
+          @chunks.each_with_index do |(kind, value), i|
+            text =
+              case kind
+              when :slot
+                field = fields[normalise(value)]
+                next unless field&.present?
+
+                field.text
+              else
+                next unless bridged?(i, present)
+
+                value
+              end
+            if split_at == i
+              out << fieldless(lead_slots, fields) if false
+              out << delim.to_s
+              split_at = nil
+            end
+            out << text unless text.nil?
+          end
+          out << delim.to_s if split_at == @chunks.length
+          terminate(out).rstrip
+        end
+
+        # 1.x segment-join semantics: present elements joined with the
+        # template's separators — a literal renders only when it
+        # separates (a run of absent elements between) two present
+        # elements; leading and trailing literals drop with the
+        # elements they would have attached to
+        def evaluate(fields)
+          present = present_indices(fields)
+          return "" if present.empty?
+
+          out = +"".dup
+          # Parenthetical groups balance on the template's literal
+          # parens alone: data text (a series run "(N.S.)", a host
+          # creator's closing paren) carries parens of its own that
+          # are no part of the template's group structure
+          paren_balance = 0
+          @chunks.each_with_index do |(kind, value), i|
+            if kind == :slot
+              field = fields[normalise(value)]
+              out << field.text if field&.present?
+            elsif value.to_s.match?(/\A[）)]/)
+              # Symmetric close: with no pending open the close-paren
+              # drops, and any trailing punctuation bridges normally
+              closes = value.to_s[/\A[）)]+/]
+              emitted = [paren_balance, closes.size].min
+              if emitted.positive?
+                out << closes[0, emitted] +
+                  value.to_s[closes.size..].to_s
+                paren_balance -= emitted
+              else
+                rest = value.to_s.sub(/\A[）)]+/, "")
+                # the trailing separator of a dropped group bridges as
+                # though the group were never there
+                if !rest.empty? && present.any? { |j| j > i } &&
+                   group_gap?(i, present)
+                  out << rest
+                end
+              end
+            elsif value.to_s.match?(/[（(]\z/)
+              # A literal opening a parenthetical group ("({{series}})",
+              # "。（{{series}}）") attaches its paren to the slot it
+              # opens — the pair drops with an absent element, while
+              # any leading punctuation bridges normally
+              lead, paren = value.to_s.match(/\A(.*)([（(])\z/).captures
+              out << lead if bridged_text?(i, lead, present)
+              if present.include?(i + 1)
+                out << paren
+                paren_balance += 1
+              end
+            elsif bridged?(i, present) ||
+                  (value.include?("<") && present.any? { |j| j < i })
+              # markup attaches to the element it closes
+              out << value
+            elsif close_pending_paren?(out, value)
+              out << value
+            end
+          end
+          terminate(out).rstrip
+        end
+
+        private
+
+        def scan(source)
+          rest = source
+          while (slot = SLOT.match(rest))
+            @chunks << [:text, rest[0...slot.begin(0)]] if
+              slot.begin(0).positive?
+            @chunks << [:slot, slot[1]]
+            rest = rest[slot.end(0)..]
+          end
+          @terminator = rest
+        end
+
+        def present_indices(fields)
+          @chunks.each_index.select do |i|
+            kind, value = @chunks[i]
+            next false unless kind == :slot
+
+            field = fields[normalise(value)]
+            !field.nil? && field.present?
+          end
+        end
+
+        # The texts between this close and the last present slot are
+        # all parenthetical-group parts (each dropped or partial)
+        def group_gap?(index, present)
+          left = present.select { |i| i < index }.max
+          @chunks[(left + 1)...index].all? do |(_, text)|
+            text.to_s.match?(/\A[）)]|[（(]\z/)
+          end
+        end
+
+        # The leading punctuation of a paren-opening literal bridges
+        # per the segment-join rule (a bare paren has no lead)
+        def bridged_text?(index, lead, present)
+          return true if lead.empty?
+
+          first_slot = @chunks.index { |(kind, _)| kind == :slot }
+          return present.include?(first_slot) if index < first_slot
+
+          left = present.select { |i| i < index }.max
+          left && present.any? { |i| i > index } &&
+            !@chunks[(left + 1)...index].any? do |(kind, _)|
+              kind == :text
+            end
+        end
+
+        # A literal is the trailing separator of the element that
+        # precedes it: it renders when that element renders and a
+        # later element renders, and it is the FIRST literal of its
+        # gap (the 1.x segment join — absent elements between two
+        # present ones leave the single separator that follows the
+        # earlier present element; an absent slot directly before a
+        # present one, as "{{date}}{{disambiguator}}, {{access}}",
+        # does not swallow it). A leading literal (no slot precedes
+        # it) attaches to the first slot, emitting only when that
+        # slot is present.
+        def bridged?(index, present)
+          first_slot = @chunks.index { |(kind, _)| kind == :slot }
+          return present.include?(first_slot) if index < first_slot
+
+          left = present.select { |i| i < index }.max
+          left && present.any? { |i| i > index } &&
+            !@chunks[(left + 1)...index].any? do |(kind, _)|
+              kind == :text
+            end
+        end
+
+        # A template's paired literal ("({{production}})") closes even
+        # when nothing follows the present element it wraps. Fullwidth
+        # parens (CJK) pair the same way
+        def close_pending_paren?(out, value)
+          return false unless value.start_with?(")", "\uFF09")
+
+          opens = out.count("(") + out.count("\uFF08")
+          closes = out.count(")") + out.count("\uFF09")
+          opens > closes
+        end
+
+        def terminate(body)
+          # 1.x punctuation cleanup: a sentence period before a spaced
+          # comma yields to the comma (", 2013" over ". , 2013"; an
+          # initials period before a comma ("P.,") stands, being
+          # unspaced), and stacked separators around absent elements
+          # collapse into one (". : ." into ". "); an empty pair left
+          # by an absent element drops
+          body = body.gsub(/:\s*\.\s*/, ": ")
+          body = body.gsub(/\.\s+,/, ",")
+          body = body.gsub(/(?:[.,:;]\s+)+[.,:;]/, ". ")
+          body = body.gsub(/\.\s*\./, ". ")
+          body = body.gsub(/:\s*\.\s*/, ": ")
+          body = body.gsub(/\(\s*\)/, "")
+          body = body.gsub(/ +/, " ").gsub(" ,", ",")
+          terminator = unbalanced_close(body, @terminator)
+          return body if terminator.strip.empty?
+          if terminator !~ /[<\w]/ && body.rstrip.end_with?(terminator.strip)
+            return body
+          end
+
+          "#{body}#{terminator}"
+        end
+
+        # The terminator's closing parens stand only when the body
+        # opened them (their "(" died with an absent element)
+        def unbalanced_close(body, terminator)
+          return terminator unless terminator.start_with?(")")
+
+          open = body.count("(") - body.count(")")
+          return terminator.sub(/\A\)+/) { |closes| closes[0, open] } if
+            open.positive?
+
+          terminator.sub(/\A\)+/, "")
+        end
+
+        def normalise(slot)
+          slot.to_s.delete("_").downcase
+        end
+      end
+    end
+  end
+end

data/lib/relaton/render/iso690.rb

@@ -0,0 +1,27 @@
+# frozen_string_literal: true
+
+module Relaton
+  module Render
+    # Citation rendering per the ISO 690 architecture: a style instance
+    # places the inventory's data elements (clause 7) in each resource
+    # kind's order (clause 8); the engine is a template evaluator over a
+    # field resolver and holds no styles, no vocabularies, no i18n.
+    # The v1 compatibility facade (relaton-render#90): isodoc's
+    # render-isodoc subclasses Relaton::Render::General.
+    autoload :General, "relaton/render/general"
+
+    module Iso690
+      autoload :Element, "relaton/render/iso690/element"
+      autoload :Elements, "relaton/render/iso690/elements"
+      autoload :Fields, "relaton/render/iso690/fields"
+      autoload :I18n, "relaton/render/iso690/i18n"
+      autoload :IndexRenderer, "relaton/render/iso690/index_renderer"
+      autoload :IdentifierKinds, "relaton/render/iso690/identifier_kinds"
+      autoload :Kinds, "relaton/render/iso690/kinds"
+      autoload :Renderer, "relaton/render/iso690/renderer"
+        autoload :Rules, "relaton/render/iso690/rules"
+      autoload :Style, "relaton/render/iso690/style"
+      autoload :Template, "relaton/render/iso690/template"
+    end
+  end
+end

data/lib/relaton/render/version.rb

@@ -1,5 +1,5 @@
 module Relaton
   module Render
-    VERSION = "0.4.2".freeze
+    VERSION = "3.0.0.pre.alpha.26".freeze
   end
 end

data/lib/relaton-render.rb

@@ -1,7 +1,13 @@
+# frozen_string_literal: true
+
 require "relaton/render/version"
-require "relaton/render/general/render"
-require "relaton/render/fields/fields"
-require "relaton/render/parse/parse"
-require "relaton/render/utils/utils"
-require "isodoc/i18n"
+require "relaton/render/iso690"
+
+module Relaton
+  module Render
+    # A bibliographic item whose style template resolves to nothing: the
+    # single-item API raises, the batch API skips
+    class Unrenderable < StandardError; end
 
+  end
+end

data/relaton-render.gemspec

@@ -19,20 +19,13 @@ Gem::Specification.new do |spec|
   spec.bindir        = "exe"
   spec.executables   = spec.files.grep(%r{^exe/}) { |f| File.basename(f) }
   spec.require_paths = ["lib"]
-  spec.required_ruby_version = Gem::Requirement.new(">= 2.5.0")
+  spec.required_ruby_version = Gem::Requirement.new(">= 3.3.0")
 
-  spec.add_development_dependency "bundler"
-  spec.add_development_dependency "debug"
-  spec.add_development_dependency "equivalent-xml", "~> 0.6"
-  spec.add_development_dependency "rake", ">= 12.3.3"
-  spec.add_development_dependency "rspec", "~> 3.0"
-  spec.add_development_dependency "simplecov"
-
-  spec.add_dependency "isodoc-i18n"
-  spec.add_dependency "liquid", "~> 4"
+  spec.add_dependency "htmlentities"
+  spec.add_development_dependency "relaton-bib"
+  spec.add_dependency "liquid", "~> 5"
+  spec.add_dependency "lutaml-model"
   spec.add_dependency "nokogiri"
-  spec.add_dependency "relaton-bib", ">= 1.12.0"
   spec.add_dependency "twitter_cldr"
-  spec.add_dependency "tzinfo-data" # we need this for windows only
-  #spec.metadata["rubygems_mfa_required"] = "true"
+  # spec.metadata["rubygems_mfa_required"] = "true"
 end