mindwords 0.3.1 → 0.5.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 6221b29c92e35c56991faead9b0065df928bb5ce5664f9511ad36a41b9382e9c
-  data.tar.gz: 9aefdce231368e36caf338567b5b8063a12e12c5468a5bd92c7395db04ff28bc
+  metadata.gz: cecbfa6443266445084268c94ccedf12ca1f1c4a8e601a510ad49d375c54e910
+  data.tar.gz: 184f031666fbc23c62ca1fb7b61bbfbce99a1bd88c75072fc9645153f4e1f502
 SHA512:
-  metadata.gz: cf7874e88c03ae3148102bf2ad8ecc539502c16bd620208303f54ea750262b9041de7c97b4c5bbf90b0aaf33a381f935b03be9228b1131be467c314bbd6b8344
-  data.tar.gz: d98408eab67356ef1312e3464ec671a381d149e7d2584ab63df6c87e4098d22c59e032dc360cd819fca2cd74d94071a5a8e400f4f87616e058f6f59a4b467c25
+  metadata.gz: 5fd523b6544be615a5a1112deca758af2cc5f6053f03e48c4382acecd413f32db9ed440d572171be342fd64141c0f34debd18f7bd6ebc615cf125029e86ce3f4
+  data.tar.gz: 047f68b8746aa20cda1bff2e9c09472fee21f70d4f36cee1599576b71f6243b5107e45f632eedf2ab2e953e9df00867dda4b9a94ec13f896f6d38a136cda9ed9

data/lib/mindwords.rb

@@ -3,6 +3,7 @@
 # file: mindwords.rb
 
 require 'rexle'
+require 'rxfhelper'
 require 'line-tree'
 
 module HashCopy
@@ -19,97 +20,15 @@ class MindWords
   using ColouredText
   using HashCopy
   
+  attr_accessor :lines
   
-  def initialize(s, parent: nil, debug: false)
+  def initialize(raws, parent: nil, debug: false)
 
-    @debug = debug
-    @a = s.strip.gsub(/^\n/,'').lines
-    @a.shift if @a.first =~ /<\?mindwords\?>/
+    @parent, @debug = parent, debug
     
-    lines = @a.map do |line|
-
-      word = line.split(/ (?=#)/,2)
-
-    end
-    
-    h = {}
-    @hashtags = {}
-
-    lines.each do |title, rawtags|
-
-      rawtags.scan(/#(\w+)/).flatten(1).each do |rawtag|
-        tag = rawtag.gsub(/ +/, '_')
-        h[tag] ||= []
-        h[tag] << title
-      end
-
-    end
-    
-    @hashtags = h.deep_clone.sort.map {|tag, fields| [tag, fields.sort]}.to_h
-
-    # does the key exist as a field? Nesting of hash object is performed here.
-
-    h.keys.each do |key|
-
-      r = h.detect {|_, value| value.include? key}
-      next unless r
-      h[r.first].delete key
-      h[r.first] << {key => h[key]}
-      #puts r.inspect
-      h.delete key
-
-    end
-
-    @h = h
-    
-    a = rexlize(h)
-    doc = Rexle.new(['root', {}, '', *a])
-
-    
-    # remove duplicates which appear in the same branch above the nested node
-    rm_duplicates(doc)
-    
-    # remove redundant nodes (outsiders)
-    # a redundant node is where all children exist in existing nested nodes
-
-    redundants = doc.root.elements.map do |e|
-
-      r = e.elements.all? {|x| !x.has_elements?}
-      puts "%s %s" % [e.name, r] if @debug
-      dups = e.elements.all? {|x| doc.root.xpath('//' + x.name).length > 1}
-      puts 'dups: ' + dups.inspect if @debug
-      e.backtrack.to_s if dups
-
-    end
-
-    redundants.compact.each {|x| doc.element(x).delete }
-    rm_duplicates(doc)
-    
-    node = if parent then
-      found = doc.root.element('//' + parent)
-      found ? found : doc.root
-    else
-      doc.root
-    end
-    
-    @outline = treeize node
-    
-    node.root.each_recursive do |e|
-    
-      e.attributes[:id] = e.attributes[:title].downcase.gsub(/ +/,'-')      
-      
-      s = e.parent.attributes[:breadcrumb] ? \
-          e.parent.attributes[:breadcrumb].to_s  + ' / ' : ''
-      e.attributes[:breadcrumb] = s +  e.value.strip
-      
-      r = @a.grep(/^#{e.attributes[:title]} #/i)
-      next unless r.any?
-      e.attributes[:hashtags] = r[0].scan(/(?<=#)\w+/).join(' ')
-
-      
-    end
-    
-    @xml = node.xml pretty: true
+    s, _ = RXFHelper.read raws
+    @lines = s.strip.gsub(/^\n/,'').lines
+    @lines.shift if @lines.first =~ /<\?mindwords\?>/
     
   end
   
@@ -119,11 +38,12 @@ class MindWords
     e =  doc.root.element("//*[@id='#{id}']")
     #e.attributes[:breadcrumb].to_s if e
     
-  end  
+  end
+
   
   def search(keyword)
     
-    a = @a.grep(/#{keyword}/i).map do |line|
+    a = @lines.grep(/#{keyword}/i).map do |line|
       
       puts 'line: ' + line.inspect if @debug
       
@@ -145,7 +65,7 @@ class MindWords
   end
   
   def sort()
-    s = @a.sort.join
+    s = @lines.sort.join
     
     def s.to_s()
       self.lines.map do |x|
@@ -158,13 +78,13 @@ class MindWords
   end
   
   def sort!()
-    @a = sort().lines
+    @lines = sort().lines
     self
   end
   
   def tag_sort()
     
-    h = @a.group_by  {|x| x[/#\w+/]}    
+    h = @lines.group_by  {|x| x[/#\w+/]}    
     s = h.sort.map {|key, value| value.sort }.join
     
     def s.to_s()
@@ -179,7 +99,7 @@ class MindWords
   end
   
   def tag_sort!()
-    @a = tag_sort().lines
+    @lines = tag_sort().lines
     self
   end
   
@@ -192,15 +112,16 @@ class MindWords
   end
 
   def to_outline(sort: true)
+    build()
     sort ? a2tree(tree_sort(LineTree.new(@outline).to_a)) : @outline
   end
   
   def to_s(colour: false)
     
     header = "<?mindwords?>\n\n"
-    return header + @a.join unless colour
+    return header + @lines.join unless colour
     
-    body = @a.map do |x|
+    body = @lines.map do |x|
         title, hashtags = x.split(/(?=#)/,2)
         title + hashtags.chomp.brown
     end.join("\n")
@@ -210,7 +131,7 @@ class MindWords
   end
   
   def to_words()
-    to_outline.lines.map {|x| x[/\w[\w ]+/] }
+    to_outline.lines.map {|x| x[/\w[\w ]+/] }.uniq
   end
   
   def to_xml()
@@ -219,6 +140,91 @@ class MindWords
 
   private
 
+  def build()
+    
+    h = {}
+
+    @lines.each do |line|
+
+      title, rawtags = line.split(/ (?=#)/,2)
+      
+      rawtags.scan(/#(\w+)/).flatten(1).each do |rawtag|
+        tag = rawtag.gsub(/ +/, '_')
+        h[tag] ||= []
+        h[tag] << title
+      end
+
+    end
+    
+    @hashtags = h.deep_clone.sort.map {|tag, fields| [tag, fields.sort]}.to_h
+
+    
+    a = rexlize(h)
+    doc = Rexle.new(['root', {}, '', *a])
+
+    # apply node nesting
+
+    doc.root.elements.each do |e|
+      
+      doc.root.xpath('//' + e.name).each do |e2|
+        
+        next if e2 === e
+        
+        e2.parent.add e
+        e2.delete
+        
+      end
+      
+    end
+
+    
+    # remove duplicates which appear in the same branch above the nested node
+    rm_duplicates(doc)
+    
+    # remove redundant nodes (outsiders)
+    # a redundant node is where all children exist in existing nested nodes
+
+    redundants = doc.root.elements.map do |e|
+
+      r = e.elements.all? {|x| !x.has_elements?}
+      puts "%s %s" % [e.name, r] if @debug
+      dups = e.elements.all? {|x| doc.root.xpath('//' + x.name).length > 1}
+      puts 'dups: ' + dups.inspect if @debug
+      e.backtrack.to_s if dups
+
+    end
+
+    redundants.compact.each {|x| doc.element(x).delete }
+ 
+    
+    node = if @parent then
+      found = doc.root.element('//' + @parent)
+      found ? found : doc.root
+    else
+      doc.root
+    end
+    
+    @outline = treeize node
+    
+    node.root.each_recursive do |e|
+    
+      e.attributes[:id] = e.attributes[:title].downcase.gsub(/ +/,'-')      
+      
+      s = e.parent.attributes[:breadcrumb] ? \
+          e.parent.attributes[:breadcrumb].to_s  + ' / ' : ''
+      e.attributes[:breadcrumb] = s +  e.value.strip
+      
+      r = @lines.grep(/^#{e.attributes[:title]} #/i)
+      next unless r.any?
+      e.attributes[:hashtags] = r[0].scan(/(?<=#)\w+/).join(' ')
+
+      
+    end
+    
+    @xml = node.xml pretty: true
+
+  end
+  
   def rexlize(a)
     
     a.map do |x|
@@ -227,16 +233,19 @@ class MindWords
 
       case x
       when String
-        [x.gsub(/ +/,'_'),  {title: x}, x]
+        [x.downcase.gsub(/ +/,''),  {title: x}, x]
       when Hash
         [
-          x.keys.first.gsub(/_/,' '), 
+          x.keys.first.downcase.gsub(/_/,' '), 
           {title: x.keys.first}, 
           x.keys.first,
          *rexlize(x.values.first)
         ]
       when Array
-        [x.first.gsub(/_/,' '), {title: x.first}, x.first, *rexlize(x.last)]
+        [
+          x.first.downcase.gsub(/_/,' '), 
+          {title: x.first}, x.first, *rexlize(x.last)
+        ]
       end
     end
 
@@ -245,24 +254,22 @@ class MindWords
   def rm_duplicates(doc)
     
     duplicates = []
+    
     doc.root.each_recursive do |e|
 
-      puts 'e: ' + e.name.inspect if @debug
       rows = e.parent.xpath('//' + e.name)
       next if rows.length < 2
 
-      rows[1..-1].each do |e2|
-        puts 'e2: ' + e2.name.inspect if @debug
-        duplicates << [e.backtrack.to_s, e2.backtrack.to_s]
-      end
+      rows[0..-2].each {|e2| duplicates << e.backtrack.to_s }
+    
     end
 
-    duplicates.each do |path, oldpath| 
-      e = doc.element(path);
-      e2 = doc.element(oldpath);  
-      next unless e2
-      e2.parent.add e
-      e2.delete
+    duplicates.each do |path|
+      
+      puts 'path: ' + path.inspect if @debug
+      e = doc.element(path)
+      e.delete if e
+      
     end   
 
   end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: mindwords
 version: !ruby/object:Gem::Version
-  version: 0.3.1
+  version: 0.5.1
 platform: ruby
 authors:
 - James Robertson
@@ -35,7 +35,7 @@ cert_chain:
   aWH7D2AmhOpqNwWnPHzWR/yzpigAVTrvpHfRxZleQj6Z/090nIH2KR0RdioMmPFq
   3+574KQzs/gR9Y5a+iMcvHRN
   -----END CERTIFICATE-----
-date: 2021-01-20 00:00:00.000000000 Z
+date: 2021-01-25 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: line-tree
@@ -57,6 +57,26 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: 0.9.1
+- !ruby/object:Gem::Dependency
+  name: rxfhelper
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.1'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 1.1.3
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.1'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 1.1.3
 description: 
 email: digital.robertson@gmail.com
 executables: []