prawn_hebrew 0.1.7 → 0.2.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 88d569ba2a0dd8db253f15e73d08a43eb1fdd8a98e6bd8aeb0976b0f90a55dfe
-  data.tar.gz: 41df57c6a0c2beccff2c7229fcc8330628e70f3cd18297701103f87e7636681f
+  metadata.gz: 07fed241f2c5f202aad8246a171d2c25d47912b0b1c99e5617e1d97228f057cc
+  data.tar.gz: 6780ab88428065941b1a9b07e9396bc26e186fb7f2d06a9c907a0402e7061162
 SHA512:
-  metadata.gz: e16f0bd9e4f962ba9e9ae9553e9ed0fcbd8fcc5bffd038490b44e059123c670a29dbd24f13a18ee105d3f743c4a2e1fc424b85dddcd0570fdad8d090c760ecd7
-  data.tar.gz: 7d47bdab8d6be871bf98932464ad82b776ff5d4cd43594ab21e07ad810dcf92e9e9d13199bb4a1f0892650c1da3279b72e7994dd908bf6c92acd21cdd5639a48
+  metadata.gz: 31a99ee31175fa68e986105b6324157e9457b9ed6bf8dad99ba3a65f9177c888b4c9284a70c7b7fe3052df041dc0c122552786ed8ea328ef44f79e2d85e243c1
+  data.tar.gz: 1d89b7a87a803d949209e0b058c49152d4b3f62eb82220a5191b08729a9206bee5b4303d886b484ce78b424aed76180563e411e35d9764d3633a721425556cdb

data/lib/prawn_hebrew.rb

@@ -2,26 +2,92 @@ require 'prawn'
 require_relative 'version'
 
 module PrawnHebrew
+  # Simple wrapper class for Hebrew text in table cells
+  class HebrewTableCell
+    attr_reader :text
+    
+    def initialize(text, hebrew_text: false)
+      @text = text
+      @hebrew_text = hebrew_text
+    end
+    
+    def to_s
+      @text.to_s
+    end
+  end
+
   module Text
     DEFAULT_HEBREW_FONT = 'GveretLevinHebrew'.freeze
     DEFAULT_ENGLISH_FONT = 'Helvetica'.freeze
     
     # Set to true for debugging which text rendering path is used
     DEBUG_MODE = false
-    INVISIBLE_CHARS = /[\u2011\u2010\u2012\u2013\u2014\u2018\u2019\u201C\u201D\u2026\u200B\u200C\u200D\u200E\u200F\uFEFF\u00AD\u202A\u202B\u202C\u202D\u202E]/.freeze
-    NBSP_CHARS = /[\u00A0\u202F]/.freeze
+    
+    # Characters that should be removed or replaced to prevent Prawn errors
+    # Zero-width and invisible characters
+    INVISIBLE_CHARS = /[\u200B\u200C\u200D\u200E\u200F\uFEFF\u00AD\u202A\u202B\u202C\u202D\u202E\u2060\u2061\u2062\u2063\u2064\u206A\u206B\u206C\u206D\u206E\u206F]/.freeze
+    
+    # Non-breaking spaces (replace with regular space)
+    NBSP_CHARS = /[\u00A0\u202F\u2007\u2008\u2009\u200A\u205F\u3000]/.freeze
+    
+    # Dashes - em dash, en dash, figure dash, horizontal bar, etc. (replace with regular hyphen)
+    DASH_CHARS = /[\u2010\u2011\u2012\u2013\u2014\u2015\u2212\uFE58\uFE63\uFF0D]/.freeze
+    
+    # Quotation marks (replace with standard quotes)
+    SMART_QUOTES_DOUBLE = /[\u201C\u201D\u201E\u201F\u00AB\u00BB\u301D\u301E\u301F]/.freeze
+    SMART_QUOTES_SINGLE = /[\u2018\u2019\u201A\u201B\u2039\u203A]/.freeze
+    
+    # Ellipsis (replace with three dots)
+    ELLIPSIS_CHAR = /\u2026/.freeze
+    
+    # Arrows and special symbols (replace with text equivalents)
+    ARROW_CHAR = /\u2192/.freeze  # → (rightward arrow)
+    LEFT_ARROW = /\u2190/.freeze  # ←
+    UP_ARROW = /\u2191/.freeze    # ↑
+    DOWN_ARROW = /\u2193/.freeze  # ↓
+    
+    # Bullet points and list markers
+    BULLET_CHARS = /[\u2022\u2023\u2043\u204C\u204D\u2219\u25E6\u25AA\u25AB\u25CF\u25CB]/.freeze
+    
+    # Other problematic characters
+    MISC_PROBLEM_CHARS = /[\u2028\u2029\uFFFC\uFFFD\uFFFF]/.freeze  # line/paragraph separators, object replacement, replacement char
+    
+    # Punctuation that should stay at the end of Hebrew text (rendered after in RTL)
+    TRAILING_PUNCTUATION = /([.,:;!?\-\u05BE\u05C3]+)$/.freeze  # includes Hebrew punctuation
+    LEADING_PUNCTUATION = /^([.,:;!?\-\u05BE\u05C3()\[\]{}]+)/.freeze
 
 
 
 
     def sanitize_text(text)
       return text if text.nil?
-      text.to_s.gsub(INVISIBLE_CHARS, ' ').gsub(NBSP_CHARS, ' ')
+      text.to_s
+        .gsub(INVISIBLE_CHARS, '')           # Remove invisible characters completely
+        .gsub(NBSP_CHARS, ' ')               # Replace non-breaking spaces with regular space
+        .gsub(DASH_CHARS, '-')               # Replace all dash types with regular hyphen
+        .gsub(SMART_QUOTES_DOUBLE, '"')      # Replace smart double quotes
+        .gsub(SMART_QUOTES_SINGLE, "'")      # Replace smart single quotes
+        .gsub(ELLIPSIS_CHAR, '...')          # Replace ellipsis with three dots
+        .gsub(ARROW_CHAR, '->')              # Replace right arrow
+        .gsub(LEFT_ARROW, '<-')              # Replace left arrow
+        .gsub(UP_ARROW, '^')                 # Replace up arrow
+        .gsub(DOWN_ARROW, 'v')               # Replace down arrow
+        .gsub(BULLET_CHARS, '*')             # Replace bullets with asterisk
+        .gsub(MISC_PROBLEM_CHARS, '')        # Remove other problematic chars
     end
 
     def hebrew_formatted_text(text, size: 12, style: :normal, hebrew_font: DEFAULT_HEBREW_FONT, english_font: DEFAULT_ENGLISH_FONT)
       text = sanitize_text(text)
       
+      # Check if text is pure Hebrew (no English characters)
+      is_pure_hebrew = text.to_s =~ /\p{Hebrew}/ && text.to_s !~ /[a-zA-Z]/
+      
+      # If pure Hebrew, render it directly without word reversal
+      if is_pure_hebrew
+        return render_pure_hebrew(text, size, style, hebrew_font)
+      end
+      
+      # Otherwise, use the mixed text logic with word reversal
       # Split by newlines first to process each line independently
       lines = text.to_s.split("\n")
       all_fragments = []
@@ -42,10 +108,8 @@ module PrawnHebrew
             hebrew_run << word
           else
             unless hebrew_run.empty?
-              hebrew_run.reverse.each_with_index do |hw, idx|
-                all_fragments << { text: hw, font: hebrew_font, size: size, direction: :rtl, styles: styles }
-                all_fragments << { text: ' ', font: hebrew_font, size: size, direction: :rtl, styles: styles } if idx < hebrew_run.length - 1
-              end
+              # Process Hebrew run and handle trailing punctuation
+              process_hebrew_run(hebrew_run, all_fragments, hebrew_font, english_font, size, styles)
               all_fragments << { text: ' ' }
               hebrew_run.clear
             end
@@ -54,10 +118,8 @@ module PrawnHebrew
         end
 
         unless hebrew_run.empty?
-          hebrew_run.reverse.each_with_index do |hw, idx|
-            all_fragments << { text: hw, font: hebrew_font, size: size, direction: :rtl, styles: styles }
-            all_fragments << { text: ' ', font: hebrew_font, size: size, direction: :rtl, styles: styles } if idx < hebrew_run.length - 1
-          end
+          # Process remaining Hebrew run
+          process_hebrew_run(hebrew_run, all_fragments, hebrew_font, english_font, size, styles)
         end
         
         # Add newline between lines (except after the last line)
@@ -68,12 +130,115 @@ module PrawnHebrew
       
       all_fragments
     end
+    
+    # Render pure Hebrew text as RTL without word reversal
+    def render_pure_hebrew(text, size, style, hebrew_font)
+      lines = text.to_s.split("\n")
+      all_fragments = []
+      
+      styles = style.is_a?(Array) ? style : [style].compact
+      
+      lines.each_with_index do |line, line_idx|
+        # For pure Hebrew, just add the line as-is with RTL direction
+        all_fragments << { text: line, font: hebrew_font, size: size, direction: :rtl, styles: styles }
+        
+        # Add newline between lines (except after the last line)
+        if line_idx < lines.length - 1
+          all_fragments << { text: "\n", font: hebrew_font, size: size, direction: :rtl, styles: styles }
+        end
+      end
+      
+      all_fragments
+    end
+    
+    # Process a run of Hebrew words, handling punctuation correctly
+    def process_hebrew_run(hebrew_run, all_fragments, hebrew_font, english_font, size, styles)
+      # Check if the last word in the run has trailing punctuation
+      last_word = hebrew_run.last
+      trailing_punct = nil
+      
+      if last_word =~ TRAILING_PUNCTUATION
+        trailing_punct = $1
+        # Remove trailing punctuation from the last word
+        hebrew_run[-1] = last_word.sub(TRAILING_PUNCTUATION, '')
+      end
+      
+      # Check if the first word has leading punctuation (for RTL, this appears at the end)
+      first_word = hebrew_run.first
+      leading_punct = nil
+      
+      if first_word =~ LEADING_PUNCTUATION
+        leading_punct = $1
+        hebrew_run[0] = first_word.sub(LEADING_PUNCTUATION, '')
+      end
+      
+      # Add leading punctuation first (it will appear at the visual end due to RTL)
+      if leading_punct
+        all_fragments << { text: leading_punct, font: english_font, size: size, styles: styles }
+      end
+      
+      # Reverse Hebrew words for RTL display
+      hebrew_run.reverse.each_with_index do |hw, idx|
+        next if hw.empty?
+        all_fragments << { text: hw, font: hebrew_font, size: size, direction: :rtl, styles: styles }
+        all_fragments << { text: ' ', font: hebrew_font, size: size, direction: :rtl, styles: styles } if idx < hebrew_run.length - 1
+      end
+      
+      # Add trailing punctuation at the end (it will appear at the visual end of Hebrew text)
+      if trailing_punct
+        all_fragments << { text: trailing_punct, font: english_font, size: size, styles: styles }
+      end
+    end
+
+    def hebrew_table(data, size: 12, style: :normal, 
+                     hebrew_font: DEFAULT_HEBREW_FONT, 
+                     english_font: DEFAULT_ENGLISH_FONT, 
+                     **table_opts)
+      # Process each row: sanitize and render Hebrew cells using formatted text
+      processed_data = data.map do |row|
+        row.map do |cell_content|
+          # Handle hash cells (e.g., {content: "text", font_style: :bold})
+          if cell_content.is_a?(Hash)
+            cell_hash = cell_content.dup
+            # Sanitize the content if it exists
+            if cell_hash[:content]
+              cell_hash[:content] = sanitize_text(cell_hash[:content].to_s)
+            end
+            cell_hash
+          else
+            # Handle simple string cells
+            sanitize_text(cell_content.to_s)
+          end
+        end
+      end
+      
+      # Create table with a block to apply Hebrew formatting to cells
+      table(processed_data, table_opts) do |table|
+        # Apply Hebrew formatting to cells that contain Hebrew text
+        table.cells.each do |cell|
+          cell_text = cell.content
+          if cell_text =~ /\p{Hebrew}/
+            # Use text_color and font to support Hebrew
+            cell.font = hebrew_font
+            cell.size = size unless cell.size  # Don't override if already set
+            cell.text_color = table_opts[:text_color] || "000000"
+          else
+            # English cells
+            cell.font = english_font
+            cell.size = size unless cell.size  # Don't override if already set
+          end
+        end
+      end
+    end
 
     def hebrew_text_box(text, size: 12, style: :normal,
                         hebrew_font: DEFAULT_HEBREW_FONT,
                         english_font: DEFAULT_ENGLISH_FONT,
                         direction: :auto, **box_opts)
       
+      # Sanitize text first to remove problematic characters
+      text = sanitize_text(text)
+      
       # Handle font specification in box_opts or use defaults
       final_hebrew_font = box_opts.delete(:hebrew_font) || hebrew_font
       final_english_font = box_opts.delete(:english_font) || english_font
@@ -223,6 +388,28 @@ module PrawnHebrew
         end
       end
     end
+
+    # Helper method to create table cells with Hebrew support
+    def hebrew_table_cell(text, size: 12, style: :normal,
+                         hebrew_font: DEFAULT_HEBREW_FONT,
+                         english_font: DEFAULT_ENGLISH_FONT,
+                         **cell_opts)
+      # Create a hash that can be used as table cell content
+      # with Hebrew text handling
+      if text.to_s =~ /\p{Hebrew}/
+        {
+          content: text,
+          hebrew_text: true,
+          size: size,
+          style: style,
+          hebrew_font: hebrew_font,
+          english_font: english_font,
+          cell_opts: cell_opts
+        }
+      else
+        { content: text }
+      end
+    end
   end
 end
 

data/lib/version.rb

@@ -1,3 +1,3 @@
 module PrawnHebrew
-  VERSION = "0.1.7"
+  VERSION = "0.2.1"
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: prawn_hebrew
 version: !ruby/object:Gem::Version
-  version: 0.1.7
+  version: 0.2.1
 platform: ruby
 authors:
 - Ben Lite