ghostwriter 1.0.1 → 1.1.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 2868e207e695355f8e9f40521b38d1f96b72b45c9ab73207396a87e5b4d535cd
-  data.tar.gz: cf111d734daa4bf94e4d9c2924dbd6c7d12b38b2b26a55db79110730f306fccc
+  metadata.gz: b87477fb4de8fc196eda6ea049eb3f1355ae583f0622c11b98543a67cfe71c8b
+  data.tar.gz: fa4e66701a3326ccceef69b3eb8f8591f8eea337f6f4f37c5b18a820bd120d42
 SHA512:
-  metadata.gz: 8d71989a44d8d2da33496172c600ec38063100c53642850982a9a93ccefbea37e40847e93edfde09d3b0f0dad98f457296f01c070387d86b11cc06e2ee9e04c1
-  data.tar.gz: 0767f0d24a895477aee922bd960380608185b741c0d87975bba65eaa03270539e2af2776bbcef2f689b604b76fde7882dbe9322e522b190858c2699359ce8a3b
+  metadata.gz: b5c545e83c7073b3c5923efc3ffcb6ecfd78650c472a482b5ec16fd7ef5533859ce6fb24d32378aafd20c83e35acd558c136856b46f0c7e0373beb7e3191be34
+  data.tar.gz: 8b259323e6653c112885ddf6fe4ee99b31c01e5625ee2fcf22d0259330dc090df11a1b7ee2dc1507c222b5f79c8322717aba9455b9bdbdca04334c0becc5959a

data/README.md

@@ -172,9 +172,9 @@ This meal we enjoy together would be improved by one.
 
 ```
 
-### Headers
+### Headings
 
-For now, headers are all treated the same and given a simple marker:
+Headings are wrapped with a marker per heading level:
 
 ```html
 <h1>Dog Maintenance and Repair</h1>
@@ -186,10 +186,12 @@ Becomes:
 
 ```
 -- Dog Maintenance and Repair --
--- Food Input Port --
--- Exhaust Port Considerations --
+---- Food Input Port ----
+------ Exhaust Port Considerations ------
 ```
 
+The `<header>` tag is treated like an `<h1>` tag.
+
 ### Lists
 
 Lists are converted, too. They are padded with newlines and are given simple markers:
@@ -211,7 +213,6 @@ Lists are converted, too. They are padded with newlines and are given simple mar
 Becomes:
 
 ```
-
 - Planes
 - Trains
 - Automobiles
@@ -219,7 +220,6 @@ Becomes:
 1. I get knocked down
 2. I get up again
 3. Never gonna keep me down
-
 ```
 
 ### Tables
@@ -265,10 +265,85 @@ Becomes:
 | Planet Express Ship | Turanga Leela   |
 ```
 
-### Presentation ARIA Role
+### Customizing Output
+
+Ghostwriter has some constructor options to customize output.
+
+You can set heading markers.
+
+```ruby
+html = <<~HTML
+   <h1>Emergency Cat Procedures</h1>
+HTML
+
+writer = Ghostwriter::Writer.new(heading_marker: '#')
+
+puts writer.textify(html)
+```
+
+Produces:
+
+```
+# Emergency Cat Procedures #
+```
+
+You can also set list item markers. Ordered markers can be anything that responds to `#next` (eg. any `Enumerator`)
+
+```ruby
+html = <<~HTML
+   <ol><li>Mercury</li><li>Venus</li><li>Mars</li></ol>
+   <ul><li>Teapot</li><li>Kettle</li></ul>
+HTML
+
+writer = Ghostwriter::Writer.new(ul_marker: '*', ol_marker: 'a')
+
+puts writer.textify(html)
+```
+
+Produces:
+
+```
+a. Mercury
+b. Venus
+c. Mars
+
+* Teapot
+* Kettle
+```
+
+And tables can be customized:
+
+```ruby
+writer = Ghostwriter::Writer.new(table_row:    '.',
+                                 table_column: '#',
+                                 table_corner: '+')
+
+puts writer.textify <<~HTML
+   <table>
+      <thead>
+         <tr><th>Moon</th><th>Portfolio</th></tr>
+      </thead>
+      <tbody>
+         <tr><td>Phobos</td><td>Fear & Panic</td></tr>
+         <tr><td>Deimos</td><td>Dread and Terror</td></tr>
+      </tbody>
+   </table>
+HTML
+```
+
+Produces:
+
+```
+# Moon   # Portfolio        #
++........+..................+
+# Phobos # Fear & Panic     #
+# Deimos # Dread and Terror #
+
+```
+
+#### Presentation ARIA Role
 
-Lists and tables with `role="presentation"` will be treated as a simple container and the normal behaviour will be
-suppressed.
+Tags with `role="presentation"` will be treated as a simple container and the normal behaviour will be suppressed.
 
 ```html
 

data/RELEASE_NOTES.md

@@ -1,5 +1,23 @@
 # Release Notes
 
+## 1.1.0 (2021-03-23)
+
+### Major
+
+* none
+
+### Minor
+
+* Added customization for headings
+* Headings now marked more for higher order headings
+* Added customization for list markers
+* Added customization for table markers
+* Writer is now immutable
+
+### Bugfixes
+
+* none
+
 ## 1.0.1 (2021-03-22)
 
 ### Major

data/lib/ghostwriter/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Ghostwriter
-   VERSION = '1.0.1'
+   VERSION = '1.1.0'
 end

data/lib/ghostwriter/writer.rb

@@ -3,12 +3,22 @@
 module Ghostwriter
    # Main Ghostwriter converter object.
    class Writer
+      attr_reader :link_base, :heading_marker, :ul_marker, :ol_marker, :table_row, :table_column, :table_corner
+
       # Creates a new ghostwriter
       #
       # @param [String] link_base the url to prefix relative links with
-      def initialize(link_base: '')
-         @link_base   = link_base
-         @list_marker = '-'
+      def initialize(link_base: '', heading_marker: '--', ul_marker: '-', ol_marker: '1',
+                     table_column: '|', table_row: '-', table_corner: '|')
+         @link_base      = link_base
+         @heading_marker = heading_marker
+         @ul_marker      = ul_marker
+         @ol_marker      = ol_marker
+         @table_column   = table_column
+         @table_row      = table_row
+         @table_corner   = table_corner
+
+         freeze
       end
 
       # Strips HTML down to plain text.
@@ -34,13 +44,13 @@ module Ghostwriter
          simple_replace(doc, 'br', "\n")
          simple_replace(doc, 'p', "\n\n")
 
-         doc.text.strip.split("\n").collect(&:strip).join("\n").concat("\n")
+         normalize_lines(doc)
       end
 
       private
 
-      def normalize_whitespace(html)
-         html.gsub(/\s/, ' ').squeeze(' ')
+      def normalize_lines(doc)
+         doc.text.strip.split("\n").collect(&:strip).join("\n").concat("\n")
       end
 
       def replace_anchors(doc)
@@ -78,8 +88,16 @@ module Ghostwriter
       end
 
       def replace_headers(doc)
-         doc.search('header, h1, h2, h3, h4, h5, h6').each do |node|
-            node.inner_html = "-- #{ node.inner_html } --\n".squeeze(' ')
+         doc.search('header, h1').each do |node|
+            node.replace("#{ @heading_marker } #{ node.inner_html } #{ @heading_marker }\n"
+                               .squeeze(' '))
+         end
+
+         (2..6).each do |n|
+            doc.search("h#{ n }").each do |node|
+               node.replace("#{ @heading_marker * n } #{ node.inner_html } #{ @heading_marker * n }\n"
+                                  .squeeze(' '))
+            end
          end
       end
 
@@ -97,41 +115,50 @@ module Ghostwriter
       end
 
       def replace_lists(doc)
-         doc.search('ul, ol').each do |list_node|
-            list_node.search('./li').each_with_index do |list_item, i|
-               marker = if list_node.node_name == 'ol'
-                           "#{ i + 1 }."
-                        else
-                           @list_marker
-                        end
-
-               list_item.inner_html = "#{ marker } #{ list_item.inner_html }\n".squeeze(' ')
-            end
+         doc.search('ol').each do |list_node|
+            replace_list_items(list_node, @ol_marker, after_marker: '.', increment: true)
+         end
 
+         doc.search('ul').each do |list_node|
+            replace_list_items(list_node, @ul_marker)
+         end
+
+         doc.search('ul, ol').each do |list_node|
             list_node.replace("#{ list_node.inner_html }\n")
          end
       end
 
+      def replace_list_items(list_node, marker, after_marker: '', increment: false)
+         list_node.search('./li').each do |list_item|
+            list_item.replace("#{ marker }#{ after_marker } #{ list_item.inner_html }\n")
+
+            marker = marker.next if increment
+         end
+      end
+
       def replace_tables(doc)
          doc.css('table').each do |table|
+            # remove whitespace between nodes
+            table.search('//text()[normalize-space()=""]').remove
+
             column_sizes = calculate_column_sizes(table)
 
             table.search('./thead/tr', './tbody/tr', './tr').each do |row|
                replace_table_nodes(row, column_sizes)
 
-               row.inner_html = "#{ row.inner_html }|\n"
+               row.replace("#{ row.inner_html }#{ @table_column }\n")
             end
 
             add_table_header_underline(table, column_sizes)
 
-            table.inner_html = "#{ table.inner_html }\n"
+            table.replace("\n#{ table.inner_html }\n")
          end
       end
 
       def calculate_column_sizes(table)
          column_sizes = table.search('tr').collect do |row|
             row.search('th', 'td').collect do |node|
-               node.inner_html.length
+               node.text.length
             end
          end
 
@@ -139,19 +166,19 @@ module Ghostwriter
       end
 
       def add_table_header_underline(table, column_sizes)
-         table.search('./thead').each do |row|
-            header_bottom = "|#{ column_sizes.collect { |len| ('-' * (len + 2)) }.join('|') }|"
+         table.search('./thead').each do |thead|
+            lines         = column_sizes.collect { |len| @table_row * (len + 2) }
+            underline_row = "#{ table_corner }#{ lines.join(@table_corner) }#{ @table_corner }"
 
-            row.inner_html = "#{ row.inner_html }#{ header_bottom }\n"
+            thead.replace("#{ thead.inner_html }#{ underline_row }\n")
          end
       end
 
       def replace_table_nodes(row, column_sizes)
          row.search('th', 'td').each_with_index do |node, i|
-            new_content = "| #{ node.inner_html }".squeeze(' ')
+            new_content = node.text.ljust(column_sizes[i] + 1)
 
-            # +2 for the extra spacing between text and pipe
-            node.inner_html = new_content.ljust(column_sizes[i] + 2)
+            node.replace("#{ @table_column } #{ new_content }")
          end
       end
 

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: ghostwriter
 version: !ruby/object:Gem::Version
-  version: 1.0.1
+  version: 1.1.0
 platform: ruby
 authors:
 - Robin Miller