pxindex-builder 0.1.2 → 0.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 67934e9d607e3570a66cf15e057f400ff5ecf589
-  data.tar.gz: fd66201c28faf4f5227ec224c81eb00ef8751a5a
+  metadata.gz: 3229af6d51962358d1f061ef1a0b97bb4c218b33
+  data.tar.gz: cef39271b8bd4e31d084c3e7cf4e1ad4b075daa2
 SHA512:
-  metadata.gz: 40f97d7d4c6a4201f63bacbaefa3374f92887285064b152f25f8d6a1f0efac63e63d390908215984ab0cc6976e24074c0daaaaa5e4eabdc26e2033cea65a42b8
-  data.tar.gz: 00c5a55685c184d0af6b9dcdbc6423e38d3120f96a4b7e48ac5014a7d3303ee02fba27512e62033343937d75f3a36569a9602e227c97f8a56f545f9ae57bbeac
+  metadata.gz: 2faa208ab98b3ade27b8d35b3efb67c0214fac5e94de2cc3e95df971495c785cc26105fd7230b075b30147893660c44dca0f6c9468a7bac8920d2e897a7be78e
+  data.tar.gz: a6cd4aea8b62250abf85e2867182536a3a7e15ba9030d16ab763729058b6f6465fa97f6045999a2c2137fb0bcd1976b3178e7415fd780ba162826b3039601e80

data/lib/pxindex-builder.rb

@@ -2,6 +2,7 @@
 
 # file: pxindex-builder.rb
 
+require 'line-tree'
 require 'wordsdotdat'
 require 'phrase_lookup'
 require 'polyrex-builder'
@@ -9,18 +10,52 @@ require 'polyrex-builder'
 
 class PxIndexBuilder
 
-  attr_reader :to_xml, :to_h
-
-  def initialize(obj, ignore=[])
-
-    h = if obj.is_a? String then
+  attr_reader :to_xml, :to_h, :to_s
+  
+  def initialize(obj, debug: false, ignore: [])
+    
+    @debug = debug
+    puts 'inside initialize: ' if @debug
+    
+    if obj.is_a? String then
     
       s, _ = RXFHelper.read(obj)
-      YAML.load(s)
+      
+      s =~ /^---/ ?  import_phrases(YAML.load(s), s, ignore) : import_index(s)
       
     elsif obj.is_a? Hash
-      obj
+      import_phrases obj, s, ignore
     end
+    
+  end
+  
+  
+  private
+  
+  def import_index(raw_s)
+    
+    # find the entries which aren't on the main index
+    s = raw_s.sub(/<[^>]+>\n/,'')
+    doc = LineTree.new(s, debug: @debug).to_doc(encapsulate: true)
+    a = doc.root.xpath('entry/text()')
+    puts 'doc: ' + doc.xml if @debug
+    a2 = doc.root.xpath('entry//entry/text()')
+    puts 'a2: ' + a2.inspect if @debug
+    a3 = a2 - a
+    puts 'a3:' + a3.inspect if @debug
+    
+    # add the new entries to the main index
+    s << "\n" + a3.join("\n")
+
+    s.prepend '<?ph schema="entries/section[heading]/entry[title, url]"?>
+
+    '
+    
+    @to_s = s
+      
+  end
+  
+  def import_phrases(h, s, ignore=[])
 
     words = h.keys.join(' ').split(/ +/).map {|x| x[/\w+/]}.uniq\
       #.tap {|x| puts 't: ' + x.inspect}
@@ -43,12 +78,13 @@ class PxIndexBuilder
     end
 
     @to_h = h = scan(index)
+    puts 'h: ' + h.inspect if @debug
     @to_xml = PolyrexBuilder.new(h, parents: %i(entry)).to_xml
 
 
   end
 
-  private
+
 
   def scan(rows)
 

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: pxindex-builder
 version: !ruby/object:Gem::Version
-  version: 0.1.2
+  version: 0.2.0
 platform: ruby
 authors:
 - James Robertson
@@ -31,8 +31,28 @@ cert_chain:
   rpKyBm8O20gPepUR+u4B3OOMZL1vUMzoAppGGdK/fa9iZHfP9tccqmgXIQ8c4X1f
   C74enOZQySX7LA==
   -----END CERTIFICATE-----
-date: 2017-12-17 00:00:00.000000000 Z
+date: 2018-04-14 00:00:00.000000000 Z
 dependencies:
+- !ruby/object:Gem::Dependency
+  name: line-tree
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.6'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 0.6.8
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.6'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 0.6.8
 - !ruby/object:Gem::Dependency
   name: wordsdotdat
   requirement: !ruby/object:Gem::Requirement
@@ -62,7 +82,7 @@ dependencies:
         version: '0.1'
     - - ">="
       - !ruby/object:Gem::Version
-        version: 0.1.5
+        version: 0.1.7
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
@@ -72,27 +92,27 @@ dependencies:
         version: '0.1'
     - - ">="
       - !ruby/object:Gem::Version
-        version: 0.1.5
+        version: 0.1.7
 - !ruby/object:Gem::Dependency
   name: polyrex-builder
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '0.2'
+        version: '0.3'
     - - ">="
       - !ruby/object:Gem::Version
-        version: 0.2.0
+        version: 0.3.0
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '0.2'
+        version: '0.3'
     - - ">="
       - !ruby/object:Gem::Version
-        version: 0.2.0
+        version: 0.3.0
 description: 
 email: james@jamesrobertson.eu
 executables: []
@@ -123,5 +143,6 @@ rubyforge_project:
 rubygems_version: 2.6.13
 signing_key: 
 specification_version: 4
-summary: Builds a Polyrex index (pxindex) XML document from a YAML document
+summary: Builds a Polyrex index (pxindex) XML document from a YAML document or an
+  indented list
 test_files: []

metadata.gz.sig

@@ -1,2 +1,2 @@
-{؃��P�`)�� ��+���*�iЮ�am�3�V<k�����N+A(������4�&߀�oq����9�ŵ�*��$d[����u	�Wq0�c/��#�!�s�6�5ΰɖD5=$v\
-.���\*�q
+_𕀣"%�Xx�6�
+sng:�R�����}�	�������W�t�XP@L�\��Ti�u�9�2Ԧ+Qf-�H���&��QK� l�m�m��$��� ?��h���f�V�U|Zz�x�������LE���(�� ���j8�E���[�A