xml_row_finder 0.2.0 → 0.3.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: e993dbcc7aa78609f6da7dad8044858a117ad8f2a35dbaffb8a4bef967d1c2fa
-  data.tar.gz: c73231bd44d7845a132c740ef02c48afc187459470e877c63372f6f82f99b03d
+  metadata.gz: 29c7aa56863bd3e1d4d7cc5bcc35c05a3c447fa6e3529b7cef763dbe2ca14091
+  data.tar.gz: 4113bc3c0d63433df71e5e04720c084a457a698eb432544c0043bc5ee4aa09f1
 SHA512:
-  metadata.gz: 996c4a63be86e7abe8948ef5dd6860f2b9226e1e429ef0f51c7ece7efb4b54cb35b18b48488403264149183b0b61d3f7d05eecc52d1dd9841f9ab0c654ffc2c6
-  data.tar.gz: fe7558dbaf9a183e860b8214bac3dfc0ca797364775c3b49ddded4c24acea51a8c600e0d3e1dc53ac74f6557dd72eb4de5284212be15bfe3df0fba7d8ca86d89
+  metadata.gz: b1424142187bb87172c1c772f54a4693c36b8f903fc20781275ea5c70003cff388472c33eb68815ea2ce1d2f70e0b0b9ba84db1caa04032615b8fc7a1a74a259
+  data.tar.gz: d34b7cf995f00cadf57c4ecec858d56540cf43ca7234b1db4937b4ac83cea440f32d6ed38479983476e37ad11e00a9d6242bae33159586981d77174992d42698

checksums.yaml.gz.sig

@@ -1,6 +1,3 @@
-�u�3h�Bju���4/+H��wt��:��H!����q�!9v��@/Ad䳹na$�]x�vw���.�r��X	�5��d����2����	���ٽR�J�O�O�:�k\��{􊒔�1��T��;��	1�
-�V��=��Ų�
-I&�3�s���Q�.V")��������&c��E+��i�͢Ԧa��eyK�ո�D~}r�PM�ߞw��h���u�{^�R?%�ۮi����
-+I�����6`��h>��
���y6l+��@�Iq=�/"���v�׵����HN�>���s4G3�/�H
-�2���?նͰ;��8O�W�߯�M�����E��
-�~���7��
+n��v,s�oF6D:�K��A���ƈiK3S���`5��h�Iᇔ����Po��
+\f���@���s�*���j"��w7�}
+���!�1uj��j^
�

data/lib/xml_row_finder.rb

@@ -2,29 +2,44 @@
 
 # file: xml_row_finder.rb
 
-require 'rexle'
-require 'rexml'
-include REXML
+require 'nokorexi'
 
 
 class XMLRowFinder
 
   attr_reader :to_a
 
-  def initialize(s, debug: false)
+  def initialize(raws, debug: false)
 
     @debug = debug
-    doc = Rexle.new(s)
-    @doc2 = Document.new(s)
+
+    @doc = if raws =~ /^http/ then
+
+      nki = Nokorexi.new(url=raws) do |doc1|
+        doc1.xpath('//*[@onclick]').each do |e|
+          e.attributes['onclick'].value = ''
+        end
+
+        doc1.xpath('//*[@onmousedown]').each do |e|
+          e.attributes['onmousedown'].value = ''
+        end
+
+      end
+
+      nki.to_doc
+
+    else
+      Rexle.new(raws)
+    end
 
     a = []
 
-    doc.root.each_recursive do |e|
+    @doc.root.each_recursive do |e|
       e.attributes.delete
       a << e.backtrack.to_xpath
     end
 
-    @to_a = a2 = a.map {|e| [a.count(e), e] }
+    @to_a = a2 = a.map {|e| [a.count(e), e] }.uniq
     xpath = a2.max_by(&:first).last
 
     a3 = xpath.split('/')
@@ -35,20 +50,60 @@ class XMLRowFinder
       p1 << a3.pop; a4 << a3.join('/') + "[%s]" % p1.reverse.join('/')
     end
 
+    # using Nokogiri since Rexle has a bug with xpath predicates
+    #
+    @doc2 = Nokogiri::XML(@doc.root.xml)
+
     a5 = a4[0..-2].map do |xpath2|
-      [XPath.match(@doc2, xpath2).length, xpath2]
+      [@doc2.xpath(xpath2).length, xpath2]
     end
 
     @xpath = a5.reverse.detect {|num, xpath2| num > 1}.last
 
+    last_row = @doc2.xpath(@xpath).last
+
+    # find the container element
+    xpath = @xpath[/^[^\[]+/]
+    axpath = xpath.split('/')
+    e = @doc.element xpath
+
+    until (e.xml.include? last_row) do
+      axpath.pop
+      e = @doc.element axpath.join('/')
+    end
+
+    @cont_xpath = axpath.join('/')
+
+  end
+
+  # returns the container element for all rows
+  # object returned: Rexle::Element
+  #
+  def body()
+    @doc.element @cont_xpath
   end
 
+  # returns the xpath pointing to the container element for all rows
+  #
+  def body_xpath()
+    @cont_xpath
+  end
+
+
+  # returns rows
+  # object returned: An array of Nokogiri XML Element object
+  #
   def rows()
-    XPath.match(@doc2, @xpath)
+    @doc2.xpath @xpath
   end
 
+  # returns the xpath pointing to the rows
+  #
   def to_xpath()
     @xpath
   end
 
+  alias rows_xpath to_xpath
+
+
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: xml_row_finder
 version: !ruby/object:Gem::Version
-  version: 0.2.0
+  version: 0.3.0
 platform: ruby
 authors:
 - James Robertson
@@ -38,25 +38,25 @@ cert_chain:
 date: 2022-01-24 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
-  name: rexle
+  name: nokorexi
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.5'
+        version: '0.5'
     - - ">="
       - !ruby/object:Gem::Version
-        version: 1.5.14
+        version: 0.5.4
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.5'
+        version: '0.5'
     - - ">="
       - !ruby/object:Gem::Version
-        version: 1.5.14
+        version: 0.5.4
 description: 
 email: digital.robertson@gmail.com
 executables: []