stationery 0.6.0 → 0.8.0

data/lib/stationery/text/entities.rb

@@ -2,23 +2,33 @@
 
 module Stationery
   module Text
-    # Decodes the HTML entities markup text may carry.
+    # Decodes the HTML character references markup text may carry: numeric
+    # (`'`, `'`, `'`) and named (the built-ins below, then the
+    # HTML 4 table). Each reference is decoded once, so `'` yields the
+    # literal text `'`; an unknown name, a malformed reference or a code
+    # point outside Unicode stays as written.
     module Entities
       NAMED = { "amp" => "&", "lt" => "<", "gt" => ">", "quot" => "\"", "apos" => "'", "nbsp" => " " }.freeze
-      PATTERN = /&(#x\h+|#\d+|[a-zA-Z]+);/
+      PATTERN = /&(#[xX]\h+|#\d+|[a-zA-Z]+);/
+      MAX_CODE_POINT = 0x10FFFF
+      SURROGATES = (0xD800..0xDFFF)
 
       module_function
 
       def decode(text)
         text.gsub(PATTERN) do |entity|
           name = Regexp.last_match(1)
-          if name.start_with?("#x") then [name[2..].to_i(16)].pack("U")
-          elsif name.start_with?("#") then [name[1..].to_i].pack("U")
+          if name.start_with?("#x", "#X") then character(name[2..].to_i(16), entity)
+          elsif name.start_with?("#") then character(name[1..].to_i, entity)
           else NAMED.fetch(name) { html4.fetch(name, entity) }
           end
         end
       end
 
+      def character(code, entity)
+        code.between?(1, MAX_CODE_POINT) && !SURROGATES.cover?(code) ? [code].pack("U") : entity
+      end
+
       def html4
         require_relative "entities/html4" unless defined?(HTML4)
         HTML4

data/lib/stationery/text/paragraph.rb

@@ -117,7 +117,7 @@ module Stationery
                                            color: style.color, letter_spacing: style.letter_spacing, rise: style.rise,
                                            opacity: style.opacity, underline: style.underline,
                                            strikethrough: style.strikethrough, kerning: style.kerning,
-                                           ligatures: style.ligatures,
+                                           ligatures: style.ligatures, features: style.features,
                                            synthetic_bold: fragment.face.synthetic_bold,
                                            synthetic_oblique: fragment.face.synthetic_oblique, word_spacing:)
         canvas.link(x, top, width, line.height, style.link, tag: link) if style.link

data/lib/stationery/text/style.rb

@@ -3,15 +3,26 @@
 module Stationery
   module Text
     SCRIPT_SCALE = 0.583
+    NO_FEATURES = [].freeze
 
     # How a run of text looks. Immutable; derive variants with #with.
+    # `hyphenate` is a bundled hyphenation language (see Hyphenation) or nil.
     Style = Data.define(:family, :size, :weight, :style, :color, :letter_spacing,
                         :underline, :strikethrough, :script, :link, :opacity, :kerning,
-                        :ligatures) do
+                        :ligatures, :features, :hyphenate) do
+      # `features:` are OpenType feature tags (`:smcp`, `"onum"`, …) applied on
+      # top of `ligatures:`; kept as sorted frozen Strings so equal sets are
+      # equal styles.
       def initialize(family:, size: 10, weight: :regular, style: :normal, color: "#000000", letter_spacing: 0,
                      underline: false, strikethrough: false, script: nil, link: nil, opacity: nil, kerning: true,
-                     ligatures: true)
-        super
+                     ligatures: true, features: NO_FEATURES, hyphenate: nil)
+        super(family:, size:, weight:, style:, color:, letter_spacing:, underline:, strikethrough:, script:, link:,
+              opacity:, kerning:, ligatures:, features: Style.features(features), hyphenate: Hyphenation.tag(hyphenate))
+      end
+
+      def self.features(tags)
+        tags = Array(tags).map(&:to_s).uniq.sort
+        tags.empty? ? NO_FEATURES : tags.freeze
       end
 
       # The size glyphs are drawn at: smaller for sub- and superscript.

data/lib/stationery/text/wrapper.rb

@@ -3,11 +3,16 @@
 module Stationery
   module Text
     # Greedy line breaking over styled runs. Breaks at spaces (the space is
-    # dropped at the break) and after hyphens; a word wider than the line is
-    # broken between characters. A word that changes style midway ("<b>Tot</b>al")
-    # is still one word.
+    # dropped at the break) and after hyphens; a word that does not fit is
+    # hyphenated when it may be (a soft hyphen U+00AD names the break points,
+    # else the style's `hyphenate` language), and a word wider than the line
+    # is broken between characters. A word that changes style midway
+    # ("<b>Tot</b>al") is still one word. Soft hyphens are never measured or
+    # drawn: a broken word draws "-" at the break and nothing otherwise.
     class Wrapper
       TOKEN = /\n|[ \t]+|[^ \t\n-]*-+|[^ \t\n-]+/
+      SOFT_HYPHEN = "­"
+      EPSILON = 0.0001
 
       Segment = Data.define(:text, :style)
 
@@ -53,16 +58,70 @@ module Stationery
         end
       end
 
-      def place_word(word)
+      # A word holding soft hyphens breaks only at them, even once the part
+      # after a break carries none (TeX's rule: a discretionary hyphen exempts
+      # the whole word from the patterns).
+      def place_word(word, explicit: word.any? { |segment| segment.text.include?(SOFT_HYPHEN) })
+        @explicit = explicit
         needed = width(@pending_space) + width(word)
-        if @current.empty? || line_width + needed <= @max + 0.0001
+        if @current.empty? || line_width + needed <= @max + EPSILON
           @current.concat(@pending_space, word)
+        elsif (head, tail = hyphenated(word, @max - line_width - width(@pending_space)))
+          @current.concat(@pending_space, head)
+          finish(wrapped: true)
+          return place_word(tail, explicit:)
         else
           finish(wrapped: true)
           @current.concat(word)
         end
         @pending_space = []
-        break_long_word while line_width > @max + 0.0001 && characters(@current) > 1
+        overflow
+      end
+
+      # A word alone on its line and still too wide: hyphenate what fits,
+      # else move the characters that overflow onto following lines.
+      def overflow
+        while line_width > @max + EPSILON && characters(@current) > 1
+          head, tail = hyphenated(@current, @max)
+          if head
+            @current = head
+            finish(wrapped: true)
+            @current = tail
+          else
+            break_long_word
+          end
+        end
+      end
+
+      # [head, tail] with the longest hyphenation of `word` whose head (and
+      # its hyphen) fits `available`; nil when the word offers no break that
+      # fits. Soft hyphens name the breaks and suppress the patterns.
+      def hyphenated(word, available)
+        chars = word.flat_map { |segment| segment.text.chars.map { |char| Segment.new(char, segment.style) } }
+        break_points(chars).reverse_each do |index|
+          head = chars.first(index).reject { |segment| segment.text == SOFT_HYPHEN }
+          head << Segment.new(hyphen_for(chars[index - 1].style), chars[index - 1].style)
+          return [head, chars.drop(index)] if width(head) <= available + EPSILON
+        end
+        nil
+      end
+
+      # Indexes into the word's characters before which it may break: after
+      # each soft hyphen, else where the language's patterns allow.
+      def break_points(chars)
+        return (1...chars.length).select { |index| chars[index - 1].text == SOFT_HYPHEN } if @explicit
+
+        text = chars.map(&:text).join
+        language = chars.first.style.hyphenate or return []
+        core = text[/\p{L}+/] or return []
+        offset = text.index(core)
+        Hyphenation.points(core, language).map { |point| point + offset }
+      end
+
+      # The font's hyphen, or the dedicated U+2010 when it lacks "-".
+      def hyphen_for(style)
+        font = @book.resolve(style).first
+        !font.glyph?("-") && font.glyph?("‐") ? "‐" : "-"
       end
 
       # Moves characters that overflow the line onto following lines.
@@ -71,7 +130,7 @@ module Stationery
         @current.each do |segment|
           segment.text.each_char do |char|
             piece = Segment.new(char, segment.style)
-            return keep_overflow(fitting, piece) if fitting.any? && width(fitting + [piece]) > @max + 0.0001
+            return keep_overflow(fitting, piece) if fitting.any? && width(fitting + [piece]) > @max + EPSILON
 
             fitting << piece
           end
@@ -104,8 +163,10 @@ module Stationery
 
       def fragments(segments)
         x = 0
-        segments.chunk_while { |a, b| a.style == b.style }.map do |group|
-          text = group.map(&:text).join
+        segments.chunk_while { |a, b| a.style == b.style }.filter_map do |group|
+          text = group.map(&:text).join.delete(SOFT_HYPHEN)
+          next if text.empty?
+
           font, face = @book.resolve(group.first.style)
           fragment_width = measure(text, group.first.style)
           Fragment.new(text, group.first.style, font, face, fragment_width, x).tap { x += fragment_width }
@@ -117,16 +178,16 @@ module Stationery
       end
 
       def line_width = width(@current)
-      def characters(segments) = segments.sum { |segment| segment.text.length }
+      def characters(segments) = segments.sum { |segment| segment.text.delete(SOFT_HYPHEN).length }
 
       def width(segments)
-        segments.sum { |segment| measure(segment.text, segment.style) }
+        segments.sum { |segment| measure(segment.text.delete(SOFT_HYPHEN), segment.style) }
       end
 
       def measure(text, style)
         font = @book.resolve(style).first
         font.width_of(text, style.render_size, letter_spacing: style.letter_spacing, kerning: style.kerning,
-                                               ligatures: style.ligatures)
+                                               ligatures: style.ligatures, features: style.features)
       end
     end
   end

data/lib/stationery/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Stationery
-  VERSION = "0.6.0"
+  VERSION = "0.8.0"
 end

data/lib/stationery/warnings.rb

@@ -29,6 +29,12 @@ module Stationery
       def message = %(image "#{source}" skipped: #{reason})
     end
 
+    OversizedImage = Data.define(:source, :pixels, :ppi, :limit) do
+      def message
+        %(image "#{source}" #{pixels}px wide is drawn at #{ppi} ppi (limit #{limit}); resize it before embedding)
+      end
+    end
+
     UnresolvedLink = Data.define(:name, :page) do
       def message = %(link to "#{name}" on page #{page} has no matching anchor)
     end
@@ -49,6 +55,10 @@ module Stationery
       def message = "tagged PDF has no language: set metadata lang:"
     end
 
+    ConformanceIssue = Data.define(:level, :subject) do
+      def message = "#{PDF::Conformance.label(level)}: #{subject} is not covered by the sRGB output intent"
+    end
+
     def initialize
       @items = []
       @glyphs = Hash.new(0)

data/lib/stationery.rb

@@ -27,6 +27,7 @@ require_relative "stationery/fonts/gpos/pair_pos"
 require_relative "stationery/fonts/kerning"
 require_relative "stationery/fonts/gsub"
 require_relative "stationery/fonts/gsub/ligature_subst"
+require_relative "stationery/fonts/gsub/single_subst"
 require_relative "stationery/fonts/ligatures"
 require_relative "stationery/fonts/subset"
 require_relative "stationery/fonts/registry"
@@ -46,6 +47,7 @@ require_relative "stationery/images/cache"
 require_relative "stationery/images/scanlines"
 require_relative "stationery/images/jpeg"
 require_relative "stationery/images/png"
+require_relative "stationery/images/resampled"
 require_relative "stationery/color"
 require_relative "stationery/geometry"
 require_relative "stationery/page"
@@ -59,11 +61,17 @@ require_relative "stationery/canvas/debug"
 require_relative "stationery/canvas/marking"
 require_relative "stationery/canvas"
 require_relative "stationery/pdf/outline_writer"
+require_relative "stationery/pdf/page_labels"
+require_relative "stationery/pdf/attachments"
+require_relative "stationery/pdf/xmp"
+require_relative "stationery/pdf/conformance"
+require_relative "stationery/pdf/factur_x"
 require_relative "stationery/forms/metrics"
 require_relative "stationery/forms/field"
 require_relative "stationery/forms/appearance"
 require_relative "stationery/forms/acro_form"
 require_relative "stationery/pdf/assembler"
+require_relative "stationery/hyphenation"
 require_relative "stationery/text/style"
 require_relative "stationery/fonts/fallback"
 require_relative "stationery/fonts/font_book"
@@ -104,6 +112,10 @@ require_relative "stationery/svg/gradient"
 require_relative "stationery/svg/shading"
 require_relative "stationery/svg/painter"
 require_relative "stationery/svg/text"
+require_relative "stationery/svg/viewport"
+require_relative "stationery/svg/region"
+require_relative "stationery/svg/clip_path"
+require_relative "stationery/svg/walker"
 require_relative "stationery/svg/document"
 require_relative "stationery/layout/svg"
 require_relative "stationery/layout/wrap"

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: stationery
 version: !ruby/object:Gem::Version
-  version: 0.6.0
+  version: 0.8.0
 platform: ruby
 authors:
 - Mikael Henriksson
@@ -71,6 +71,7 @@ files:
 - lib/stationery/fonts/gpos/pair_pos.rb
 - lib/stationery/fonts/gsub.rb
 - lib/stationery/fonts/gsub/ligature_subst.rb
+- lib/stationery/fonts/gsub/single_subst.rb
 - lib/stationery/fonts/installer.rb
 - lib/stationery/fonts/kern_table.rb
 - lib/stationery/fonts/kerning.rb
@@ -94,10 +95,17 @@ files:
 - lib/stationery/html/tree_builder/blocks.rb
 - lib/stationery/html/tree_builder/collector.rb
 - lib/stationery/html/whitespace.rb
+- lib/stationery/hyphenation.rb
+- lib/stationery/hyphenation/patterns/LICENSES.md
+- lib/stationery/hyphenation/patterns/de-1996.pat.txt
+- lib/stationery/hyphenation/patterns/en-us.hyp.txt
+- lib/stationery/hyphenation/patterns/en-us.pat.txt
+- lib/stationery/hyphenation/patterns/sv.pat.txt
 - lib/stationery/images/cache.rb
 - lib/stationery/images/images.rb
 - lib/stationery/images/jpeg.rb
 - lib/stationery/images/png.rb
+- lib/stationery/images/resampled.rb
 - lib/stationery/images/scanlines.rb
 - lib/stationery/instrumentation.rb
 - lib/stationery/layout/box.rb
@@ -137,16 +145,23 @@ files:
 - lib/stationery/page.rb
 - lib/stationery/page_templates.rb
 - lib/stationery/pdf/assembler.rb
+- lib/stationery/pdf/attachments.rb
+- lib/stationery/pdf/conformance.rb
 - lib/stationery/pdf/encryption/aes.rb
 - lib/stationery/pdf/encryption/rc4.rb
 - lib/stationery/pdf/encryption/revision4.rb
 - lib/stationery/pdf/encryption/revision6.rb
 - lib/stationery/pdf/encryption/standard_security.rb
+- lib/stationery/pdf/factur_x.rb
+- lib/stationery/pdf/icc/LICENSE.md
+- lib/stationery/pdf/icc/sRGB2014.icc
 - lib/stationery/pdf/outline_writer.rb
+- lib/stationery/pdf/page_labels.rb
 - lib/stationery/pdf/serializer.rb
 - lib/stationery/pdf/stream.rb
 - lib/stationery/pdf/types.rb
 - lib/stationery/pdf/writer.rb
+- lib/stationery/pdf/xmp.rb
 - lib/stationery/preview.rb
 - lib/stationery/rails.rb
 - lib/stationery/rails/previews_controller.rb
@@ -162,11 +177,13 @@ files:
 - lib/stationery/structure.rb
 - lib/stationery/svg/arc.rb
 - lib/stationery/svg/bounds.rb
+- lib/stationery/svg/clip_path.rb
 - lib/stationery/svg/document.rb
 - lib/stationery/svg/gradient.rb
 - lib/stationery/svg/painter.rb
 - lib/stationery/svg/parser.rb
 - lib/stationery/svg/path_data.rb
+- lib/stationery/svg/region.rb
 - lib/stationery/svg/selector.rb
 - lib/stationery/svg/shading.rb
 - lib/stationery/svg/shapes.rb
@@ -174,6 +191,8 @@ files:
 - lib/stationery/svg/stylesheet.rb
 - lib/stationery/svg/text.rb
 - lib/stationery/svg/transform.rb
+- lib/stationery/svg/viewport.rb
+- lib/stationery/svg/walker.rb
 - lib/stationery/tagging/element.rb
 - lib/stationery/tagging/tree.rb
 - lib/stationery/tagging/writer.rb