stationery 0.8.0 → 0.10.0

data/lib/stationery/text/wrapper.rb

@@ -3,7 +3,8 @@
 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 that does not fit is
+    # dropped at the break), after hyphens, at a zero-width space (U+200B,
+    # HTML's <wbr>) and between ideographic characters (see Breaks); 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
@@ -12,6 +13,7 @@ module Stationery
     class Wrapper
       TOKEN = /\n|[ \t]+|[^ \t\n-]*-+|[^ \t\n-]+/
       SOFT_HYPHEN = "­"
+      ZERO_WIDTH_SPACE = Breaks::ZERO_WIDTH_SPACE
       EPSILON = 0.0001
 
       Segment = Data.define(:text, :style)
@@ -26,6 +28,7 @@ module Stationery
         @current = []
         @pending_space = []
         @fallback = fallback_style
+        @widths = {}.compare_by_identity
         items(runs).each { |item| place(item) }
         finish unless @current.empty? && @lines.any? && !@ended_with_newline
         @lines
@@ -33,22 +36,67 @@ module Stationery
 
       private
 
-      # [:word, segments] | [:space, segments] | [:newline]
+      # [:word, segments] | [:space, segments] | [:newline]. A word that runs
+      # on into the next styled run stays one word ("<b>Tot</b>al"); a
+      # zero-width space is an empty :space, a break that draws nothing; a
+      # word with CJK characters is cut into its break units (Breaks).
       def items(runs)
         items = []
+        @join = false # whether the next token may continue the last word
         runs.each do |run|
-          run.text.scan(TOKEN) do |token|
-            if token == "\n" then items << [:newline]
-            elsif token.match?(/\A[ \t]/) then items << [:space, [Segment.new(token, run.style)]]
-            elsif items.last&.first == :word && !items.last.last.last.text.end_with?("-")
-              items.last.last << Segment.new(token, run.style)
-            else items << [:word, [Segment.new(token, run.style)]]
+          if run.text.include?(ZERO_WIDTH_SPACE)
+            run.text.split(ZERO_WIDTH_SPACE, -1).each_with_index do |part, index|
+              if index.positive?
+                items << [:space, []]
+                @join = false
+              end
+              tokens(items, part, run.style)
             end
+          else
+            tokens(items, run.text, run.style)
           end
         end
         items
       end
 
+      def tokens(items, text, style)
+        text.scan(TOKEN) do |token|
+          if token == "\n"
+            items << [:newline]
+            @join = false
+          elsif token.match?(/\A[ \t]/)
+            items << [:space, [Segment.new(token, style)]]
+            @join = false
+          else
+            word_units(items, token, style)
+          end
+        end
+      end
+
+      # A Latin word is one unit and may run on into the next styled run
+      # unless it ends in a hyphen; a word with CJK characters is cut into
+      # its break units, none of which the next token continues. Latin words
+      # skip the unit split, so they cost no allocation of their own.
+      def word_units(items, token, style)
+        if Breaks.cjk?(token)
+          Breaks.units(token).each do |text, glued|
+            add_unit(items, text, style, glued)
+            @join = false
+          end
+        else
+          add_unit(items, token, style, false)
+          @join = !token.end_with?("-")
+        end
+      end
+
+      def add_unit(items, text, style, glued)
+        if (@join || glued) && items.last&.first == :word
+          items.last.last << Segment.new(text, style)
+        else
+          items << [:word, [Segment.new(text, style)]]
+        end
+      end
+
       def place((kind, segments))
         @ended_with_newline = kind == :newline
         case kind
@@ -97,6 +145,8 @@ module Stationery
       # 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)
+        return unless @explicit || word.first&.style&.hyphenate
+
         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 }
@@ -112,6 +162,8 @@ module Stationery
         return (1...chars.length).select { |index| chars[index - 1].text == SOFT_HYPHEN } if @explicit
 
         text = chars.map(&:text).join
+        return [] if Breaks.cjk?(text)
+
         language = chars.first.style.hyphenate or return []
         core = text[/\p{L}+/] or return []
         offset = text.index(core)
@@ -164,7 +216,7 @@ module Stationery
       def fragments(segments)
         x = 0
         segments.chunk_while { |a, b| a.style == b.style }.filter_map do |group|
-          text = group.map(&:text).join.delete(SOFT_HYPHEN)
+          text = plain(group.map(&:text).join)
           next if text.empty?
 
           font, face = @book.resolve(group.first.style)
@@ -178,12 +230,18 @@ module Stationery
       end
 
       def line_width = width(@current)
-      def characters(segments) = segments.sum { |segment| segment.text.delete(SOFT_HYPHEN).length }
+      def characters(segments) = segments.sum { |segment| plain(segment.text).length }
 
+      # A line is summed again for every word that joins it; its segments
+      # are the same objects each time, so each is measured once per wrap.
       def width(segments)
-        segments.sum { |segment| measure(segment.text.delete(SOFT_HYPHEN), segment.style) }
+        segments.sum { |segment| @widths[segment] ||= measure(plain(segment.text), segment.style) }
       end
 
+      # The text without its soft hyphens; the same String when it has none,
+      # which is nearly always, so measuring a word allocates nothing.
+      def plain(text) = text.include?(SOFT_HYPHEN) ? text.delete(SOFT_HYPHEN) : text
+
       def measure(text, style)
         font = @book.resolve(style).first
         font.width_of(text, style.render_size, letter_spacing: style.letter_spacing, kerning: style.kerning,

data/lib/stationery/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Stationery
-  VERSION = "0.8.0"
+  VERSION = "0.10.0"
 end

data/lib/stationery/warnings.rb

@@ -43,6 +43,19 @@ module Stationery
       def message = %(link "#{href}" dropped: scheme not allowed)
     end
 
+    NestingLimit = Data.define(:depth, :limit) do
+      def message = "content nested #{depth} levels deep was flattened below level #{limit}"
+    end
+
+    UnsupportedCss = Data.define(:properties, :selectors) do
+      def message
+        parts = []
+        parts << "properties #{properties.join(", ")}" if properties.any?
+        parts << "selectors #{selectors.join(", ")}" if selectors.any?
+        "html styles not read: #{parts.join("; ")}"
+      end
+    end
+
     DuplicateAnchor = Data.define(:name, :page) do
       def message = %(anchor "#{name}" on page #{page} is already defined)
     end

data/lib/stationery.rb

@@ -66,7 +66,11 @@ require_relative "stationery/pdf/attachments"
 require_relative "stationery/pdf/xmp"
 require_relative "stationery/pdf/conformance"
 require_relative "stationery/pdf/factur_x"
+require_relative "stationery/pdf/signature"
+require_relative "stationery/pdf/signature/cms"
+require_relative "stationery/pdf/signature/timestamp"
 require_relative "stationery/forms/metrics"
+require_relative "stationery/forms/typeface"
 require_relative "stationery/forms/field"
 require_relative "stationery/forms/appearance"
 require_relative "stationery/forms/acro_form"
@@ -79,6 +83,7 @@ require_relative "stationery/text/entities"
 require_relative "stationery/text/markup"
 require_relative "stationery/text/runs_builder"
 require_relative "stationery/text/line"
+require_relative "stationery/text/breaks"
 require_relative "stationery/text/wrapper"
 require_relative "stationery/text/paragraph"
 require_relative "stationery/layout/node"
@@ -103,8 +108,7 @@ require_relative "stationery/svg/path_data"
 require_relative "stationery/svg/arc"
 require_relative "stationery/svg/parser"
 require_relative "stationery/svg/transform"
-require_relative "stationery/svg/selector"
-require_relative "stationery/svg/stylesheet"
+require_relative "stationery/css"
 require_relative "stationery/svg/style"
 require_relative "stationery/svg/shapes"
 require_relative "stationery/svg/bounds"

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: stationery
 version: !ruby/object:Gem::Version
-  version: 0.8.0
+  version: 0.10.0
 platform: ruby
 authors:
 - Mikael Henriksson
@@ -40,6 +40,9 @@ files:
 - lib/stationery/cli/render.rb
 - lib/stationery/color.rb
 - lib/stationery/component.rb
+- lib/stationery/css.rb
+- lib/stationery/css/selector.rb
+- lib/stationery/css/stylesheet.rb
 - lib/stationery/document.rb
 - lib/stationery/elements.rb
 - lib/stationery/elements/forms.rb
@@ -88,7 +91,9 @@ files:
 - lib/stationery/forms/appearance.rb
 - lib/stationery/forms/field.rb
 - lib/stationery/forms/metrics.rb
+- lib/stationery/forms/typeface.rb
 - lib/stationery/geometry.rb
+- lib/stationery/html/css.rb
 - lib/stationery/html/document.rb
 - lib/stationery/html/tokenizer.rb
 - lib/stationery/html/tree_builder.rb
@@ -158,6 +163,9 @@ files:
 - lib/stationery/pdf/outline_writer.rb
 - lib/stationery/pdf/page_labels.rb
 - lib/stationery/pdf/serializer.rb
+- lib/stationery/pdf/signature.rb
+- lib/stationery/pdf/signature/cms.rb
+- lib/stationery/pdf/signature/timestamp.rb
 - lib/stationery/pdf/stream.rb
 - lib/stationery/pdf/types.rb
 - lib/stationery/pdf/writer.rb
@@ -168,8 +176,11 @@ files:
 - lib/stationery/railtie.rb
 - lib/stationery/regions.rb
 - lib/stationery/resources.rb
+- lib/stationery/rich/nesting.rb
 - lib/stationery/rich/nodes.rb
 - lib/stationery/rich/renderer.rb
+- lib/stationery/rich/renderer/boxes.rb
+- lib/stationery/rich/renderer/indents.rb
 - lib/stationery/rich/renderer/inlines.rb
 - lib/stationery/rich/renderer/links.rb
 - lib/stationery/rich/styles.rb
@@ -184,11 +195,9 @@ files:
 - 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
 - lib/stationery/svg/style.rb
-- lib/stationery/svg/stylesheet.rb
 - lib/stationery/svg/text.rb
 - lib/stationery/svg/transform.rb
 - lib/stationery/svg/viewport.rb
@@ -200,6 +209,7 @@ files:
 - lib/stationery/testing/marked_text.rb
 - lib/stationery/testing/matchers.rb
 - lib/stationery/testing/structure_reader.rb
+- lib/stationery/text/breaks.rb
 - lib/stationery/text/entities.rb
 - lib/stationery/text/entities/html4.rb
 - lib/stationery/text/line.rb