ae_easy-text 0.0.1 → 0.0.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 239189344e783f67b085da7394e535aa693a4b067c62b8d0b16f733a0b19d4f7
-  data.tar.gz: ca144105f26e399116b05560ff870f6aa051a04696602f6f68f67f06b9e0bfda
+  metadata.gz: 10296214e4de01abc2d77f5a5549c9e4c883009a86915908283bb71dad3bec0b
+  data.tar.gz: 9db77d0892191a3dd5c170ffe2257c14559f0721075c88c1fc75ce6fa0f0b04e
 SHA512:
-  metadata.gz: 0b7c4495eeb71e5dae3ad799d14f8a2d83989a949183ee3df2837191b4a4f3a10965ead38416ccda078da7b29fc083eb02fd53a24999f97473b02f77489d921c
-  data.tar.gz: 4f377b26bcfb0ef4cce7806d153fb97de115d0e0bc4beef5e43b55b0125e1936d6c4440d1131cbe9cb4d5fe28a1810c2820269f6c317511068c91aff42ad8126
+  metadata.gz: 6582399f34051ebcc5fa5192c22fc984329a04b1ffefb8c2db49ac5e782978b49c3fc4ea5c7fe6c0f0bbdffb4fff1799adb03f79ca92358b13409691e10b6b0d
+  data.tar.gz: 174c5533dd32772393fb7a8ba632fb4de8fe5ce8eba6e35d753abc906b295c1659d6cde0cf18e7496b6469a430e7f922335e23f7c3d5d724eba019163a6801f3

data/ae_easy-text.gemspec

@@ -38,7 +38,7 @@ Gem::Specification.new do |spec|
   spec.require_paths = ["lib"]
   spec.required_ruby_version = '>= 2.2.2'
 
-  spec.add_dependency 'ae_easy-core', '>= 0'
+  spec.add_dependency 'ae_easy-core', '>= 0.1.2'
   spec.add_development_dependency 'bundler', '>= 1.16.3'
   spec.add_development_dependency 'rake', '>= 10.0'
   spec.add_development_dependency 'minitest', '>= 5.11'

data/doc/AeEasy.html

@@ -107,7 +107,7 @@
 </div>
 
       <div id="footer">
-  Generated on Tue Feb 26 16:50:02 2019 by
+  Generated on Fri Mar  8 17:26:54 2019 by
   <a href="http://yardoc.org" title="Yay! A Ruby Documentation Tool" target="_parent">yard</a>
   0.9.18 (ruby-2.5.3).
 </div>

data/doc/AeEasy/Text.html

@@ -108,7 +108,7 @@
 
 </div>
         </dt>
-        <dd><pre class="code"><span class='tstring'><span class='tstring_beg'>&quot;</span><span class='tstring_content'>0.0.1</span><span class='tstring_end'>&quot;</span></span></pre></dd>
+        <dd><pre class="code"><span class='tstring'><span class='tstring_beg'>&quot;</span><span class='tstring_content'>0.0.2</span><span class='tstring_end'>&quot;</span></span></pre></dd>
       
     </dl>
   
@@ -326,7 +326,7 @@ using a header map to match columns.</p>
         <li class="public ">
   <span class="summary_signature">
     
-      <a href="#strip-class_method" title="strip (class method)">.<strong>strip</strong>(raw_text)  &#x21d2; String<sup>?</sup> </a>
+      <a href="#strip-class_method" title="strip (class method)">.<strong>strip</strong>(raw_text, orig_encoding = &#39;ASCII&#39;)  &#x21d2; String<sup>?</sup> </a>
     
 
     
@@ -341,7 +341,8 @@ using a header map to match columns.</p>
 
   
     <span class="summary_desc"><div class='inline'>
-<p>Strip a value.</p>
+<p>Strip a value by trimming spaces, reducing secuential spaces into a  
+single space, decode HTML entities and change encoding to UTF-8.</p>
 </div></span>
   
 </li>
@@ -532,17 +533,19 @@ using a header map to match columns.</p>
       <pre class="lines">
 
 
-60
-61
 62
-63</pre>
+63
+64
+65
+66</pre>
     </td>
     <td>
-      <pre class="code"><span class="info file"># File 'lib/ae_easy/text.rb', line 60</span>
+      <pre class="code"><span class="info file"># File 'lib/ae_easy/text.rb', line 62</span>
 
 <span class='kw'>def</span> <span class='kw'>self</span><span class='period'>.</span><span class='id identifier rubyid_default_parser'>default_parser</span> <span class='id identifier rubyid_cell_element'>cell_element</span><span class='comma'>,</span> <span class='id identifier rubyid_data'>data</span><span class='comma'>,</span> <span class='id identifier rubyid_key'>key</span>
-  <span class='id identifier rubyid_cell_element'>cell_element</span><span class='op'>&amp;.</span><span class='id identifier rubyid_search'>search</span><span class='lparen'>(</span><span class='tstring'><span class='tstring_beg'>&#39;</span><span class='tstring_content'>//i</span><span class='tstring_end'>&#39;</span></span><span class='rparen'>)</span><span class='period'>.</span><span class='id identifier rubyid_remove'>remove</span>
-  <span class='id identifier rubyid_row_data'>row_data</span><span class='lbracket'>[</span><span class='id identifier rubyid_key'>key</span><span class='rbracket'>]</span> <span class='op'>=</span> <span class='id identifier rubyid_strip'>strip</span> <span class='id identifier rubyid_cell_element'>cell_element</span><span class='op'>&amp;.</span><span class='id identifier rubyid_text'>text</span>
+  <span class='kw'>return</span> <span class='kw'>if</span> <span class='id identifier rubyid_cell_element'>cell_element</span><span class='period'>.</span><span class='id identifier rubyid_nil?'>nil?</span>
+  <span class='id identifier rubyid_cell_element'>cell_element</span><span class='period'>.</span><span class='id identifier rubyid_search'>search</span><span class='lparen'>(</span><span class='tstring'><span class='tstring_beg'>&#39;</span><span class='tstring_content'>//i</span><span class='tstring_end'>&#39;</span></span><span class='rparen'>)</span><span class='period'>.</span><span class='id identifier rubyid_remove'>remove</span> <span class='kw'>if</span> <span class='id identifier rubyid_cell_element'>cell_element</span><span class='period'>.</span><span class='id identifier rubyid_search'>search</span><span class='lparen'>(</span><span class='tstring'><span class='tstring_beg'>&#39;</span><span class='tstring_content'>//i</span><span class='tstring_end'>&#39;</span></span><span class='rparen'>)</span><span class='period'>.</span><span class='id identifier rubyid_count'>count</span> <span class='op'>&gt;</span> <span class='int'>0</span>
+  <span class='id identifier rubyid_data'>data</span><span class='lbracket'>[</span><span class='id identifier rubyid_key'>key</span><span class='rbracket'>]</span> <span class='op'>=</span> <span class='id identifier rubyid_strip'>strip</span> <span class='id identifier rubyid_cell_element'>cell_element</span><span class='period'>.</span><span class='id identifier rubyid_text'>text</span>
 <span class='kw'>end</span></pre>
     </td>
   </tr>
@@ -815,6 +818,22 @@ ignored.</p>
           
             &mdash; <div class='inline'>
 <p>Custom column parsers for advance data extraction.</p>
+</div>
+          
+        </li>
+      
+        <li>
+          <span class="name">:ignore_text_nodes</span>
+          <span class="type">(<tt>Boolean</tt>)</span>
+          <span class="default">
+            
+              &mdash; default:
+              <tt>true</tt>
+            
+          </span>
+          
+            &mdash; <div class='inline'>
+<p>Ignore text nodes when retriving content cells and rows.</p>
 </div>
           
         </li>
@@ -917,11 +936,6 @@ ignored.</p>
       <pre class="lines">
 
 
-84
-85
-86
-87
-88
 89
 90
 91
@@ -955,10 +969,21 @@ ignored.</p>
 119
 120
 121
-122</pre>
+122
+123
+124
+125
+126
+127
+128
+129
+130
+131
+132
+133</pre>
     </td>
     <td>
-      <pre class="code"><span class="info file"># File 'lib/ae_easy/text.rb', line 84</span>
+      <pre class="code"><span class="info file"># File 'lib/ae_easy/text.rb', line 89</span>
 
 <span class='kw'>def</span> <span class='kw'>self</span><span class='period'>.</span><span class='id identifier rubyid_parse_content'>parse_content</span> <span class='id identifier rubyid_opts'>opts</span><span class='comma'>,</span> <span class='op'>&amp;</span><span class='id identifier rubyid_filter'>filter</span>
   <span class='id identifier rubyid_opts'>opts</span> <span class='op'>=</span> <span class='lbrace'>{</span>
@@ -966,7 +991,8 @@ ignored.</p>
     <span class='label'>selector:</span> <span class='kw'>nil</span><span class='comma'>,</span>
     <span class='label'>first_row_header:</span> <span class='kw'>false</span><span class='comma'>,</span>
     <span class='label'>header_map:</span> <span class='lbrace'>{</span><span class='rbrace'>}</span><span class='comma'>,</span>
-    <span class='label'>column_parsers:</span> <span class='lbrace'>{</span><span class='rbrace'>}</span>
+    <span class='label'>column_parsers:</span> <span class='lbrace'>{</span><span class='rbrace'>}</span><span class='comma'>,</span>
+    <span class='label'>ignore_text_nodes:</span> <span class='kw'>true</span>
   <span class='rbrace'>}</span><span class='period'>.</span><span class='id identifier rubyid_merge'>merge</span> <span class='id identifier rubyid_opts'>opts</span>
 
   <span class='comment'># Setup config
@@ -975,10 +1001,13 @@ ignored.</p>
   <span class='id identifier rubyid_first'>first</span> <span class='op'>=</span> <span class='id identifier rubyid_first_row_header'>first_row_header</span> <span class='op'>=</span> <span class='id identifier rubyid_opts'>opts</span><span class='lbracket'>[</span><span class='symbol'>:first_row_header</span><span class='rbracket'>]</span>
   <span class='id identifier rubyid_header_map'>header_map</span> <span class='op'>=</span> <span class='id identifier rubyid_opts'>opts</span><span class='lbracket'>[</span><span class='symbol'>:header_map</span><span class='rbracket'>]</span>
   <span class='id identifier rubyid_column_parsers'>column_parsers</span> <span class='op'>=</span> <span class='id identifier rubyid_opts'>opts</span><span class='lbracket'>[</span><span class='symbol'>:column_parsers</span><span class='rbracket'>]</span>
+  <span class='id identifier rubyid_ignore_text_nodes'>ignore_text_nodes</span> <span class='op'>=</span> <span class='id identifier rubyid_opts'>opts</span><span class='lbracket'>[</span><span class='symbol'>:ignore_text_nodes</span><span class='rbracket'>]</span>
 
   <span class='comment'># Get and parse rows
 </span>  <span class='id identifier rubyid_html_rows'>html_rows</span> <span class='op'>=</span> <span class='id identifier rubyid_opts'>opts</span><span class='lbracket'>[</span><span class='symbol'>:html</span><span class='rbracket'>]</span><span class='period'>.</span><span class='id identifier rubyid_css'>css</span><span class='lparen'>(</span><span class='id identifier rubyid_opts'>opts</span><span class='lbracket'>[</span><span class='symbol'>:selector</span><span class='rbracket'>]</span><span class='rparen'>)</span>
   <span class='id identifier rubyid_html_rows'>html_rows</span><span class='period'>.</span><span class='id identifier rubyid_each'>each</span> <span class='kw'>do</span> <span class='op'>|</span><span class='id identifier rubyid_row'>row</span><span class='op'>|</span>
+    <span class='kw'>next</span> <span class='kw'>if</span> <span class='id identifier rubyid_ignore_text_nodes'>ignore_text_nodes</span> <span class='op'>&amp;&amp;</span> <span class='id identifier rubyid_row'>row</span><span class='period'>.</span><span class='id identifier rubyid_name'>name</span> <span class='op'>==</span> <span class='tstring'><span class='tstring_beg'>&#39;</span><span class='tstring_content'>text</span><span class='tstring_end'>&#39;</span></span>
+
     <span class='comment'># First row header validation
 </span>    <span class='kw'>if</span> <span class='id identifier rubyid_first'>first</span> <span class='op'>&amp;&amp;</span> <span class='id identifier rubyid_first_row_header'>first_row_header</span>
       <span class='id identifier rubyid_first'>first</span> <span class='op'>=</span> <span class='kw'>false</span>
@@ -989,7 +1018,9 @@ ignored.</p>
 </span>    <span class='id identifier rubyid_row_data'>row_data</span> <span class='op'>=</span> <span class='lbrace'>{</span><span class='rbrace'>}</span>
     <span class='id identifier rubyid_header_map'>header_map</span><span class='period'>.</span><span class='id identifier rubyid_each'>each</span> <span class='kw'>do</span> <span class='op'>|</span><span class='id identifier rubyid_key'>key</span><span class='comma'>,</span> <span class='id identifier rubyid_index'>index</span><span class='op'>|</span>
       <span class='comment'># Parse column html with default or custom parser
-</span>      <span class='id identifier rubyid_child_element'>child_element</span> <span class='op'>=</span> <span class='id identifier rubyid_row'>row</span><span class='period'>.</span><span class='id identifier rubyid_children'>children</span><span class='lbracket'>[</span><span class='id identifier rubyid_index'>index</span><span class='rbracket'>]</span>
+</span>      <span class='id identifier rubyid_children'>children</span> <span class='op'>=</span> <span class='id identifier rubyid_row'>row</span><span class='period'>.</span><span class='id identifier rubyid_children'>children</span>
+      <span class='id identifier rubyid_children'>children</span> <span class='op'>=</span> <span class='id identifier rubyid_children'>children</span><span class='period'>.</span><span class='id identifier rubyid_select'>select</span><span class='lbrace'>{</span><span class='op'>|</span><span class='id identifier rubyid_i'>i</span><span class='op'>|</span><span class='id identifier rubyid_i'>i</span><span class='period'>.</span><span class='id identifier rubyid_name'>name</span> <span class='op'>!=</span> <span class='tstring'><span class='tstring_beg'>&#39;</span><span class='tstring_content'>text</span><span class='tstring_end'>&#39;</span></span><span class='rbrace'>}</span> <span class='kw'>if</span> <span class='id identifier rubyid_ignore_text_nodes'>ignore_text_nodes</span>
+      <span class='id identifier rubyid_child_element'>child_element</span> <span class='op'>=</span> <span class='id identifier rubyid_children'>children</span><span class='lbracket'>[</span><span class='id identifier rubyid_index'>index</span><span class='rbracket'>]</span>
       <span class='id identifier rubyid_column_parsers'>column_parsers</span><span class='lbracket'>[</span><span class='id identifier rubyid_key'>key</span><span class='rbracket'>]</span><span class='period'>.</span><span class='id identifier rubyid_nil?'>nil?</span> <span class='op'>?</span>
         <span class='id identifier rubyid_default_parser'>default_parser</span><span class='lparen'>(</span><span class='id identifier rubyid_child_element'>child_element</span><span class='comma'>,</span> <span class='id identifier rubyid_row_data'>row_data</span><span class='comma'>,</span> <span class='id identifier rubyid_key'>key</span><span class='rparen'>)</span> <span class='op'>:</span>
         <span class='id identifier rubyid_column_parsers'>column_parsers</span><span class='lbracket'>[</span><span class='id identifier rubyid_key'>key</span><span class='rbracket'>]</span><span class='period'>.</span><span class='id identifier rubyid_call'>call</span><span class='lparen'>(</span><span class='id identifier rubyid_child_element'>child_element</span><span class='comma'>,</span> <span class='id identifier rubyid_row_data'>row_data</span><span class='comma'>,</span> <span class='id identifier rubyid_key'>key</span><span class='rparen'>)</span>
@@ -1106,6 +1137,22 @@ ignored.</p>
             &mdash; <div class='inline'>
 <p>If true then selector first matching row will be used as header for
 parsing.</p>
+</div>
+          
+        </li>
+      
+        <li>
+          <span class="name">:ignore_text_nodes</span>
+          <span class="type">(<tt>Boolean</tt>)</span>
+          <span class="default">
+            
+              &mdash; default:
+              <tt>true</tt>
+            
+          </span>
+          
+            &mdash; <div class='inline'>
+<p>Ignore text nodes when retriving header cells and rows.</p>
 </div>
           
         </li>
@@ -1138,20 +1185,6 @@ parsing.</p>
       <pre class="lines">
 
 
-152
-153
-154
-155
-156
-157
-158
-159
-160
-161
-162
-163
-164
-165
 166
 167
 168
@@ -1166,21 +1199,43 @@ parsing.</p>
 177
 178
 179
-180</pre>
+180
+181
+182
+183
+184
+185
+186
+187
+188
+189
+190
+191
+192
+193
+194
+195
+196
+197
+198
+199
+200</pre>
     </td>
     <td>
-      <pre class="code"><span class="info file"># File 'lib/ae_easy/text.rb', line 152</span>
+      <pre class="code"><span class="info file"># File 'lib/ae_easy/text.rb', line 166</span>
 
 <span class='kw'>def</span> <span class='kw'>self</span><span class='period'>.</span><span class='id identifier rubyid_parse_header_map'>parse_header_map</span> <span class='id identifier rubyid_opts'>opts</span> <span class='op'>=</span> <span class='lbrace'>{</span><span class='rbrace'>}</span>
   <span class='id identifier rubyid_opts'>opts</span> <span class='op'>=</span> <span class='lbrace'>{</span>
     <span class='label'>html:</span> <span class='kw'>nil</span><span class='comma'>,</span>
     <span class='label'>selector:</span> <span class='kw'>nil</span><span class='comma'>,</span>
     <span class='label'>column_key_label_map:</span> <span class='lbrace'>{</span><span class='rbrace'>}</span><span class='comma'>,</span>
-    <span class='label'>first_row_header:</span> <span class='kw'>false</span>
+    <span class='label'>first_row_header:</span> <span class='kw'>false</span><span class='comma'>,</span>
+    <span class='label'>ignore_text_nodes:</span> <span class='kw'>true</span>
   <span class='rbrace'>}</span><span class='period'>.</span><span class='id identifier rubyid_merge'>merge</span> <span class='id identifier rubyid_opts'>opts</span>
 
   <span class='comment'># Setup config
 </span>  <span class='id identifier rubyid_dictionary'>dictionary</span> <span class='op'>=</span> <span class='id identifier rubyid_opts'>opts</span><span class='lbracket'>[</span><span class='symbol'>:column_key_label_map</span><span class='rbracket'>]</span>
+  <span class='id identifier rubyid_ignore_text_nodes'>ignore_text_nodes</span> <span class='op'>=</span> <span class='id identifier rubyid_opts'>opts</span><span class='lbracket'>[</span><span class='symbol'>:ignore_text_nodes</span><span class='rbracket'>]</span>
   <span class='id identifier rubyid_data'>data</span> <span class='op'>=</span> <span class='lbracket'>[</span><span class='rbracket'>]</span>
   <span class='id identifier rubyid_column_map'>column_map</span> <span class='op'>=</span> <span class='kw'>nil</span>
 
@@ -1189,8 +1244,12 @@ parsing.</p>
   <span class='kw'>return</span> <span class='kw'>nil</span> <span class='kw'>if</span> <span class='id identifier rubyid_html_rows'>html_rows</span><span class='period'>.</span><span class='id identifier rubyid_nil?'>nil?</span>
   <span class='id identifier rubyid_html_rows'>html_rows</span> <span class='op'>=</span> <span class='lbracket'>[</span><span class='id identifier rubyid_html_rows'>html_rows</span><span class='period'>.</span><span class='id identifier rubyid_first'>first</span><span class='rbracket'>]</span> <span class='kw'>if</span> <span class='id identifier rubyid_opts'>opts</span><span class='lbracket'>[</span><span class='symbol'>:first_row_header</span><span class='rbracket'>]</span>
   <span class='id identifier rubyid_html_rows'>html_rows</span><span class='period'>.</span><span class='id identifier rubyid_each'>each</span> <span class='kw'>do</span> <span class='op'>|</span><span class='id identifier rubyid_row'>row</span><span class='op'>|</span>
+    <span class='kw'>next</span> <span class='kw'>if</span> <span class='id identifier rubyid_ignore_text_nodes'>ignore_text_nodes</span> <span class='op'>&amp;&amp;</span> <span class='id identifier rubyid_row'>row</span><span class='period'>.</span><span class='id identifier rubyid_name'>name</span> <span class='op'>==</span> <span class='tstring'><span class='tstring_beg'>&#39;</span><span class='tstring_content'>text</span><span class='tstring_end'>&#39;</span></span>
+
     <span class='id identifier rubyid_column_map'>column_map</span> <span class='op'>=</span> <span class='lbrace'>{</span><span class='rbrace'>}</span>
-    <span class='id identifier rubyid_row'>row</span><span class='period'>.</span><span class='id identifier rubyid_children'>children</span><span class='period'>.</span><span class='id identifier rubyid_each_with_index'>each_with_index</span> <span class='kw'>do</span> <span class='op'>|</span><span class='id identifier rubyid_col'>col</span><span class='comma'>,</span> <span class='id identifier rubyid_index'>index</span><span class='op'>|</span>
+    <span class='id identifier rubyid_children'>children</span> <span class='op'>=</span> <span class='id identifier rubyid_row'>row</span><span class='period'>.</span><span class='id identifier rubyid_children'>children</span>
+    <span class='id identifier rubyid_children'>children</span> <span class='op'>=</span> <span class='id identifier rubyid_children'>children</span><span class='period'>.</span><span class='id identifier rubyid_select'>select</span><span class='lbrace'>{</span><span class='op'>|</span><span class='id identifier rubyid_i'>i</span><span class='op'>|</span><span class='id identifier rubyid_i'>i</span><span class='period'>.</span><span class='id identifier rubyid_name'>name</span> <span class='op'>!=</span> <span class='tstring'><span class='tstring_beg'>&#39;</span><span class='tstring_content'>text</span><span class='tstring_end'>&#39;</span></span><span class='rbrace'>}</span> <span class='kw'>if</span> <span class='id identifier rubyid_ignore_text_nodes'>ignore_text_nodes</span>
+    <span class='id identifier rubyid_children'>children</span><span class='period'>.</span><span class='id identifier rubyid_each_with_index'>each_with_index</span> <span class='kw'>do</span> <span class='op'>|</span><span class='id identifier rubyid_col'>col</span><span class='comma'>,</span> <span class='id identifier rubyid_index'>index</span><span class='op'>|</span>
       <span class='comment'># Parse and map column header
 </span>      <span class='id identifier rubyid_column_key'>column_key</span> <span class='op'>=</span> <span class='id identifier rubyid_translate_label_to_key'>translate_label_to_key</span> <span class='id identifier rubyid_col'>col</span><span class='comma'>,</span> <span class='id identifier rubyid_dictionary'>dictionary</span>
       <span class='kw'>next</span> <span class='kw'>if</span> <span class='id identifier rubyid_column_key'>column_key</span><span class='period'>.</span><span class='id identifier rubyid_nil?'>nil?</span>
@@ -1336,6 +1395,22 @@ parsing.</p>
           
             &mdash; <div class='inline'>
 <p>Custom column parsers for advance data extraction.</p>
+</div>
+          
+        </li>
+      
+        <li>
+          <span class="name">:ignore_text_nodes</span>
+          <span class="type">(<tt>Boolean</tt>)</span>
+          <span class="default">
+            
+              &mdash; default:
+              <tt>true</tt>
+            
+          </span>
+          
+            &mdash; <div class='inline'>
+<p>Ignore text nodes when retriving cells and rows.</p>
 </div>
           
         </li>
@@ -1443,32 +1518,35 @@ parsing.</p>
       <pre class="lines">
 
 
-204
-205
-206
-207
-208
-209
-210
-211
-212
-213
-214
-215
-216
-217
-218
-219
-220
-221
-222
-223
-224
-225
-226</pre>
+226
+227
+228
+229
+230
+231
+232
+233
+234
+235
+236
+237
+238
+239
+240
+241
+242
+243
+244
+245
+246
+247
+248
+249
+250
+251</pre>
     </td>
     <td>
-      <pre class="code"><span class="info file"># File 'lib/ae_easy/text.rb', line 204</span>
+      <pre class="code"><span class="info file"># File 'lib/ae_easy/text.rb', line 226</span>
 
 <span class='kw'>def</span> <span class='kw'>self</span><span class='period'>.</span><span class='id identifier rubyid_parse_table'>parse_table</span> <span class='id identifier rubyid_opts'>opts</span> <span class='op'>=</span> <span class='lbrace'>{</span><span class='rbrace'>}</span><span class='comma'>,</span> <span class='op'>&amp;</span><span class='id identifier rubyid_filter'>filter</span>
   <span class='id identifier rubyid_opts'>opts</span> <span class='op'>=</span> <span class='lbrace'>{</span>
@@ -1477,19 +1555,22 @@ parsing.</p>
     <span class='label'>header_key_label_map:</span> <span class='lbrace'>{</span><span class='rbrace'>}</span><span class='comma'>,</span>
     <span class='label'>content_selector:</span> <span class='kw'>nil</span><span class='comma'>,</span>
     <span class='label'>first_row_header:</span> <span class='kw'>false</span><span class='comma'>,</span>
-    <span class='label'>column_parsers:</span> <span class='lbrace'>{</span><span class='rbrace'>}</span>
+    <span class='label'>column_parsers:</span> <span class='lbrace'>{</span><span class='rbrace'>}</span><span class='comma'>,</span>
+    <span class='label'>ignore_text_nodes:</span> <span class='kw'>true</span>
   <span class='rbrace'>}</span><span class='period'>.</span><span class='id identifier rubyid_merge'>merge</span> <span class='id identifier rubyid_opts'>opts</span>
   <span class='kw'>return</span> <span class='kw'>nil</span> <span class='kw'>if</span> <span class='id identifier rubyid_opts'>opts</span><span class='lbracket'>[</span><span class='symbol'>:html</span><span class='rbracket'>]</span><span class='period'>.</span><span class='id identifier rubyid_nil?'>nil?</span>
   <span class='id identifier rubyid_header_map'>header_map</span> <span class='op'>=</span> <span class='kw'>self</span><span class='period'>.</span><span class='id identifier rubyid_parse_header_map'>parse_header_map</span> <span class='label'>html:</span> <span class='id identifier rubyid_opts'>opts</span><span class='lbracket'>[</span><span class='symbol'>:html</span><span class='rbracket'>]</span><span class='comma'>,</span>
     <span class='label'>selector:</span> <span class='id identifier rubyid_opts'>opts</span><span class='lbracket'>[</span><span class='symbol'>:header_selector</span><span class='rbracket'>]</span><span class='comma'>,</span>
     <span class='label'>column_key_label_map:</span> <span class='id identifier rubyid_opts'>opts</span><span class='lbracket'>[</span><span class='symbol'>:header_key_label_map</span><span class='rbracket'>]</span><span class='comma'>,</span>
-    <span class='label'>first_row_header:</span> <span class='id identifier rubyid_opts'>opts</span><span class='lbracket'>[</span><span class='symbol'>:first_row_header</span><span class='rbracket'>]</span>
+    <span class='label'>first_row_header:</span> <span class='id identifier rubyid_opts'>opts</span><span class='lbracket'>[</span><span class='symbol'>:first_row_header</span><span class='rbracket'>]</span><span class='comma'>,</span>
+    <span class='label'>ignore_text_nodes:</span> <span class='id identifier rubyid_opts'>opts</span><span class='lbracket'>[</span><span class='symbol'>:ignore_text_nodes</span><span class='rbracket'>]</span>
   <span class='kw'>return</span> <span class='kw'>nil</span> <span class='kw'>if</span> <span class='id identifier rubyid_header_map'>header_map</span><span class='period'>.</span><span class='id identifier rubyid_nil?'>nil?</span>
   <span class='id identifier rubyid_data'>data</span> <span class='op'>=</span> <span class='kw'>self</span><span class='period'>.</span><span class='id identifier rubyid_parse_content'>parse_content</span> <span class='label'>html:</span> <span class='id identifier rubyid_opts'>opts</span><span class='lbracket'>[</span><span class='symbol'>:html</span><span class='rbracket'>]</span><span class='comma'>,</span>
     <span class='label'>selector:</span> <span class='id identifier rubyid_opts'>opts</span><span class='lbracket'>[</span><span class='symbol'>:content_selector</span><span class='rbracket'>]</span><span class='comma'>,</span>
     <span class='label'>header_map:</span> <span class='id identifier rubyid_header_map'>header_map</span><span class='comma'>,</span>
     <span class='label'>first_row_header:</span> <span class='id identifier rubyid_opts'>opts</span><span class='lbracket'>[</span><span class='symbol'>:first_row_header</span><span class='rbracket'>]</span><span class='comma'>,</span>
     <span class='label'>column_parsers:</span> <span class='id identifier rubyid_opts'>opts</span><span class='lbracket'>[</span><span class='symbol'>:column_parsers</span><span class='rbracket'>]</span><span class='comma'>,</span>
+    <span class='label'>ignore_text_nodes:</span> <span class='id identifier rubyid_opts'>opts</span><span class='lbracket'>[</span><span class='symbol'>:ignore_text_nodes</span><span class='rbracket'>]</span><span class='comma'>,</span>
     <span class='op'>&amp;</span><span class='id identifier rubyid_filter'>filter</span>
   <span class='lbrace'>{</span><span class='label'>header_map:</span> <span class='id identifier rubyid_header_map'>header_map</span><span class='comma'>,</span> <span class='label'>data:</span> <span class='id identifier rubyid_data'>data</span><span class='rbrace'>}</span>
 <span class='kw'>end</span></pre>
@@ -1625,6 +1706,22 @@ parsing.</p>
           
             &mdash; <div class='inline'>
 <p>Custom column parsers for advance data extraction.</p>
+</div>
+          
+        </li>
+      
+        <li>
+          <span class="name">:ignore_text_nodes</span>
+          <span class="type">(<tt>Boolean</tt>)</span>
+          <span class="default">
+            
+              &mdash; default:
+              <tt>true</tt>
+            
+          </span>
+          
+            &mdash; <div class='inline'>
+<p>Ignore text nodes when retriving cells and rows.</p>
 </div>
           
         </li>
@@ -1732,42 +1829,43 @@ parsing.</p>
       <pre class="lines">
 
 
-249
-250
-251
-252
-253
-254
-255
-256
-257
-258
-259
-260
-261
-262
-263
-264
-265
-266
-267
-268
-269
-270
-271
-272
-273
-274
-275
 276
 277
 278
 279
 280
-281</pre>
+281
+282
+283
+284
+285
+286
+287
+288
+289
+290
+291
+292
+293
+294
+295
+296
+297
+298
+299
+300
+301
+302
+303
+304
+305
+306
+307
+308
+309</pre>
     </td>
     <td>
-      <pre class="code"><span class="info file"># File 'lib/ae_easy/text.rb', line 249</span>
+      <pre class="code"><span class="info file"># File 'lib/ae_easy/text.rb', line 276</span>
 
 <span class='kw'>def</span> <span class='kw'>self</span><span class='period'>.</span><span class='id identifier rubyid_parse_vertical_table'>parse_vertical_table</span> <span class='id identifier rubyid_opts'>opts</span> <span class='op'>=</span> <span class='lbrace'>{</span><span class='rbrace'>}</span><span class='comma'>,</span> <span class='op'>&amp;</span><span class='id identifier rubyid_filter'>filter</span>
   <span class='id identifier rubyid_opts'>opts</span> <span class='op'>=</span> <span class='lbrace'>{</span>
@@ -1776,7 +1874,8 @@ parsing.</p>
     <span class='label'>header_selector:</span> <span class='kw'>nil</span><span class='comma'>,</span>
     <span class='label'>header_key_label_map:</span> <span class='lbrace'>{</span><span class='rbrace'>}</span><span class='comma'>,</span>
     <span class='label'>content_selector:</span> <span class='kw'>nil</span><span class='comma'>,</span>
-    <span class='label'>column_parsers:</span> <span class='lbrace'>{</span><span class='rbrace'>}</span>
+    <span class='label'>column_parsers:</span> <span class='lbrace'>{</span><span class='rbrace'>}</span><span class='comma'>,</span>
+    <span class='label'>ignore_text_nodes:</span> <span class='kw'>true</span>
   <span class='rbrace'>}</span><span class='period'>.</span><span class='id identifier rubyid_merge'>merge</span> <span class='id identifier rubyid_opts'>opts</span>
   <span class='kw'>return</span> <span class='kw'>nil</span> <span class='kw'>if</span> <span class='id identifier rubyid_opts'>opts</span><span class='lbracket'>[</span><span class='symbol'>:html</span><span class='rbracket'>]</span><span class='period'>.</span><span class='id identifier rubyid_nil?'>nil?</span>
 
@@ -1810,7 +1909,7 @@ parsing.</p>
       <div class="method_details ">
   <h3 class="signature " id="strip-class_method">
   
-    .<strong>strip</strong>(raw_text)  &#x21d2; <tt>String</tt><sup>?</sup> 
+    .<strong>strip</strong>(raw_text, orig_encoding = &#39;ASCII&#39;)  &#x21d2; <tt>String</tt><sup>?</sup> 
   
 
   
@@ -1819,7 +1918,10 @@ parsing.</p>
 </h3><div class="docstring">
   <div class="discussion">
     
-<p>Strip a value.</p>
+<p>Strip a value by trimming spaces, reducing secuential spaces into a</p>
+
+<pre class="code ruby"><code class="ruby">single space, decode HTML entities and change encoding to UTF-8.
+</code></pre>
 
 
   </div>
@@ -1840,6 +1942,24 @@ parsing.</p>
         &mdash;
         <div class='inline'>
 <p>Text to strip.</p>
+</div>
+      
+    </li>
+  
+    <li>
+      
+        <span class='name'>orig_encoding</span>
+      
+      
+        <span class='type'>(<tt>String</tt>)</span>
+      
+      
+        <em class="default">(defaults to: <tt>&#39;ASCII&#39;</tt>)</em>
+      
+      
+        &mdash;
+        <div class='inline'>
+<p>Text original encoding.</p>
 </div>
       
     </li>
@@ -1871,8 +1991,6 @@ parsing.</p>
       <pre class="lines">
 
 
-42
-43
 44
 45
 46
@@ -1882,21 +2000,23 @@ parsing.</p>
 50
 51
 52
-53</pre>
+53
+54
+55</pre>
     </td>
     <td>
-      <pre class="code"><span class="info file"># File 'lib/ae_easy/text.rb', line 42</span>
+      <pre class="code"><span class="info file"># File 'lib/ae_easy/text.rb', line 44</span>
 
-<span class='kw'>def</span> <span class='kw'>self</span><span class='period'>.</span><span class='id identifier rubyid_strip'>strip</span> <span class='id identifier rubyid_raw_text'>raw_text</span>
+<span class='kw'>def</span> <span class='kw'>self</span><span class='period'>.</span><span class='id identifier rubyid_strip'>strip</span> <span class='id identifier rubyid_raw_text'>raw_text</span><span class='comma'>,</span> <span class='id identifier rubyid_orig_encoding'>orig_encoding</span> <span class='op'>=</span> <span class='tstring'><span class='tstring_beg'>&#39;</span><span class='tstring_content'>ASCII</span><span class='tstring_end'>&#39;</span></span>
   <span class='kw'>return</span> <span class='kw'>nil</span> <span class='kw'>if</span> <span class='id identifier rubyid_raw_text'>raw_text</span><span class='period'>.</span><span class='id identifier rubyid_nil?'>nil?</span>
   <span class='id identifier rubyid_raw_text'>raw_text</span> <span class='op'>=</span> <span class='id identifier rubyid_raw_text'>raw_text</span><span class='period'>.</span><span class='id identifier rubyid_to_s'>to_s</span> <span class='kw'>unless</span> <span class='id identifier rubyid_raw_text'>raw_text</span><span class='period'>.</span><span class='id identifier rubyid_is_a?'>is_a?</span> <span class='const'>String</span>
   <span class='id identifier rubyid_regex'>regex</span> <span class='op'>=</span> <span class='tstring'><span class='regexp_beg'>/</span><span class='tstring_content'>(\s|\u3000|\u00a0)+</span><span class='regexp_end'>/</span></span>
   <span class='id identifier rubyid_good_encoding'>good_encoding</span> <span class='op'>=</span> <span class='lparen'>(</span><span class='id identifier rubyid_raw_text'>raw_text</span> <span class='op'>=~</span> <span class='tstring'><span class='regexp_beg'>/</span><span class='tstring_content'>\u3000</span><span class='regexp_end'>/</span></span> <span class='op'>||</span> <span class='kw'>true</span><span class='rparen'>)</span> <span class='kw'>rescue</span> <span class='kw'>false</span>
   <span class='kw'>unless</span> <span class='id identifier rubyid_good_encoding'>good_encoding</span>
-    <span class='id identifier rubyid_raw_text'>raw_text</span> <span class='op'>=</span> <span class='id identifier rubyid_raw_text'>raw_text</span><span class='period'>.</span><span class='id identifier rubyid_force_encoding'>force_encoding</span><span class='lparen'>(</span><span class='gvar'>$APP_CONFIG</span><span class='lbracket'>[</span><span class='symbol'>:encoding</span><span class='rbracket'>]</span><span class='rparen'>)</span><span class='period'>.</span><span class='id identifier rubyid_encode'>encode</span><span class='lparen'>(</span><span class='tstring'><span class='tstring_beg'>&#39;</span><span class='tstring_content'>UTF-8</span><span class='tstring_end'>&#39;</span></span><span class='rparen'>)</span>
+    <span class='id identifier rubyid_raw_text'>raw_text</span> <span class='op'>=</span> <span class='id identifier rubyid_raw_text'>raw_text</span><span class='period'>.</span><span class='id identifier rubyid_force_encoding'>force_encoding</span><span class='lparen'>(</span><span class='id identifier rubyid_orig_encoding'>orig_encoding</span><span class='rparen'>)</span><span class='period'>.</span><span class='id identifier rubyid_encode'>encode</span><span class='lparen'>(</span><span class='tstring'><span class='tstring_beg'>&#39;</span><span class='tstring_content'>UTF-8</span><span class='tstring_end'>&#39;</span></span><span class='comma'>,</span> <span class='label'>invalid:</span> <span class='symbol'>:replace</span><span class='comma'>,</span> <span class='label'>undef:</span> <span class='symbol'>:replace</span><span class='rparen'>)</span>
     <span class='id identifier rubyid_regex'>regex</span> <span class='op'>=</span> <span class='tstring'><span class='regexp_beg'>/</span><span class='tstring_content'>(\s|\u3000|\u00a0|\u00c2\u00a0)+</span><span class='regexp_end'>/</span></span>
   <span class='kw'>end</span>
-  <span class='id identifier rubyid_text'>text</span> <span class='op'>=</span> <span class='id identifier rubyid_raw_text'>raw_text</span><span class='op'>&amp;.</span><span class='id identifier rubyid_gsub'>gsub</span><span class='lparen'>(</span><span class='id identifier rubyid_regex'>regex</span><span class='comma'>,</span> <span class='tstring'><span class='tstring_beg'>&#39;</span><span class='tstring_content'> </span><span class='tstring_end'>&#39;</span></span><span class='rparen'>)</span><span class='op'>&amp;.</span><span class='id identifier rubyid_strip'>strip</span>
+  <span class='id identifier rubyid_text'>text</span> <span class='op'>=</span> <span class='id identifier rubyid_raw_text'>raw_text</span><span class='period'>.</span><span class='id identifier rubyid_gsub'>gsub</span><span class='lparen'>(</span><span class='id identifier rubyid_regex'>regex</span><span class='comma'>,</span> <span class='tstring'><span class='tstring_beg'>&#39;</span><span class='tstring_content'> </span><span class='tstring_end'>&#39;</span></span><span class='rparen'>)</span><span class='period'>.</span><span class='id identifier rubyid_strip'>strip</span>
   <span class='id identifier rubyid_text'>text</span><span class='period'>.</span><span class='id identifier rubyid_nil?'>nil?</span> <span class='op'>?</span> <span class='kw'>nil</span> <span class='op'>:</span> <span class='id identifier rubyid_decode_html'>decode_html</span><span class='lparen'>(</span><span class='id identifier rubyid_text'>text</span><span class='rparen'>)</span>
 <span class='kw'>end</span></pre>
     </td>
@@ -1984,25 +2104,27 @@ parsing.</p>
       <pre class="lines">
 
 
-131
-132
-133
-134
-135
-136
-137
-138</pre>
+142
+143
+144
+145
+146
+147
+148
+149
+150</pre>
     </td>
     <td>
-      <pre class="code"><span class="info file"># File 'lib/ae_easy/text.rb', line 131</span>
+      <pre class="code"><span class="info file"># File 'lib/ae_easy/text.rb', line 142</span>
 
 <span class='kw'>def</span> <span class='kw'>self</span><span class='period'>.</span><span class='id identifier rubyid_translate_label_to_key'>translate_label_to_key</span> <span class='id identifier rubyid_element'>element</span><span class='comma'>,</span> <span class='id identifier rubyid_label_map'>label_map</span>
-  <span class='id identifier rubyid_element'>element</span><span class='op'>&amp;.</span><span class='id identifier rubyid_search'>search</span><span class='lparen'>(</span><span class='tstring'><span class='tstring_beg'>&#39;</span><span class='tstring_content'>//i</span><span class='tstring_end'>&#39;</span></span><span class='rparen'>)</span><span class='period'>.</span><span class='id identifier rubyid_remove'>remove</span>
-  <span class='id identifier rubyid_text'>text</span> <span class='op'>=</span> <span class='id identifier rubyid_strip'>strip</span> <span class='id identifier rubyid_element'>element</span><span class='op'>&amp;.</span><span class='id identifier rubyid_text'>text</span>
-  <span class='id identifier rubyid_key'>key</span> <span class='op'>=</span> <span class='id identifier rubyid_label_map'>label_map</span><span class='period'>.</span><span class='id identifier rubyid_find'>find</span> <span class='kw'>do</span> <span class='op'>|</span><span class='id identifier rubyid_k'>k</span><span class='comma'>,</span><span class='id identifier rubyid_v'>v</span><span class='op'>|</span>
+  <span class='kw'>return</span> <span class='kw'>nil</span> <span class='kw'>if</span> <span class='id identifier rubyid_element'>element</span><span class='period'>.</span><span class='id identifier rubyid_nil?'>nil?</span>
+  <span class='id identifier rubyid_element'>element</span><span class='period'>.</span><span class='id identifier rubyid_search'>search</span><span class='lparen'>(</span><span class='tstring'><span class='tstring_beg'>&#39;</span><span class='tstring_content'>//i</span><span class='tstring_end'>&#39;</span></span><span class='rparen'>)</span><span class='period'>.</span><span class='id identifier rubyid_remove'>remove</span> <span class='kw'>if</span> <span class='id identifier rubyid_element'>element</span><span class='period'>.</span><span class='id identifier rubyid_search'>search</span><span class='lparen'>(</span><span class='tstring'><span class='tstring_beg'>&#39;</span><span class='tstring_content'>//i</span><span class='tstring_end'>&#39;</span></span><span class='rparen'>)</span><span class='period'>.</span><span class='id identifier rubyid_count'>count</span> <span class='op'>&gt;</span> <span class='int'>0</span>
+  <span class='id identifier rubyid_text'>text</span> <span class='op'>=</span> <span class='id identifier rubyid_strip'>strip</span> <span class='id identifier rubyid_element'>element</span><span class='period'>.</span><span class='id identifier rubyid_text'>text</span>
+  <span class='id identifier rubyid_key_pair'>key_pair</span> <span class='op'>=</span> <span class='id identifier rubyid_label_map'>label_map</span><span class='period'>.</span><span class='id identifier rubyid_find'>find</span> <span class='kw'>do</span> <span class='op'>|</span><span class='id identifier rubyid_k'>k</span><span class='comma'>,</span><span class='id identifier rubyid_v'>v</span><span class='op'>|</span>
     <span class='id identifier rubyid_v'>v</span><span class='period'>.</span><span class='id identifier rubyid_is_a?'>is_a?</span><span class='lparen'>(</span><span class='const'>Regexp</span><span class='rparen'>)</span> <span class='op'>?</span> <span class='lparen'>(</span><span class='id identifier rubyid_text'>text</span> <span class='op'>=~</span> <span class='id identifier rubyid_v'>v</span><span class='rparen'>)</span> <span class='op'>:</span> <span class='lparen'>(</span><span class='id identifier rubyid_text'>text</span> <span class='op'>==</span> <span class='id identifier rubyid_v'>v</span><span class='rparen'>)</span>
-  <span class='kw'>end</span><span class='op'>&amp;.</span><span class='id identifier rubyid_first'>first</span>
-  <span class='id identifier rubyid_key'>key</span>
+  <span class='kw'>end</span>
+  <span class='id identifier rubyid_key'>key</span> <span class='op'>=</span> <span class='id identifier rubyid_key_pair'>key_pair</span><span class='period'>.</span><span class='id identifier rubyid_nil?'>nil?</span> <span class='op'>?</span> <span class='kw'>nil</span> <span class='op'>:</span> <span class='id identifier rubyid_key_pair'>key_pair</span><span class='lbracket'>[</span><span class='int'>0</span><span class='rbracket'>]</span>
 <span class='kw'>end</span></pre>
     </td>
   </tr>
@@ -2014,7 +2136,7 @@ parsing.</p>
 </div>
 
       <div id="footer">
-  Generated on Tue Feb 26 16:50:03 2019 by
+  Generated on Fri Mar  8 17:26:54 2019 by
   <a href="http://yardoc.org" title="Yay! A Ruby Documentation Tool" target="_parent">yard</a>
   0.9.18 (ruby-2.5.3).
 </div>

data/doc/_index.html

@@ -112,7 +112,7 @@
 </div>
 
       <div id="footer">
-  Generated on Tue Feb 26 16:50:02 2019 by
+  Generated on Fri Mar  8 17:26:54 2019 by
   <a href="http://yardoc.org" title="Yay! A Ruby Documentation Tool" target="_parent">yard</a>
   0.9.18 (ruby-2.5.3).
 </div>

data/doc/file.README.html

@@ -81,7 +81,7 @@ href="http://rubydoc.org/gems/ae_easy-text/frames">here</a>.</p>
 </div></div>
 
       <div id="footer">
-  Generated on Tue Feb 26 16:50:02 2019 by
+  Generated on Fri Mar  8 17:26:54 2019 by
   <a href="http://yardoc.org" title="Yay! A Ruby Documentation Tool" target="_parent">yard</a>
   0.9.18 (ruby-2.5.3).
 </div>

data/doc/index.html

@@ -81,7 +81,7 @@ href="http://rubydoc.org/gems/ae_easy-text/frames">here</a>.</p>
 </div></div>
 
       <div id="footer">
-  Generated on Tue Feb 26 16:50:02 2019 by
+  Generated on Fri Mar  8 17:26:54 2019 by
   <a href="http://yardoc.org" title="Yay! A Ruby Documentation Tool" target="_parent">yard</a>
   0.9.18 (ruby-2.5.3).
 </div>

data/doc/top-level-namespace.html

@@ -100,7 +100,7 @@
 </div>
 
       <div id="footer">
-  Generated on Tue Feb 26 16:50:02 2019 by
+  Generated on Fri Mar  8 17:26:54 2019 by
   <a href="http://yardoc.org" title="Yay! A Ruby Documentation Tool" target="_parent">yard</a>
   0.9.18 (ruby-2.5.3).
 </div>

data/lib/ae_easy/text.rb

@@ -1,7 +1,7 @@
 require 'cgi'
 require 'json'
 require 'digest/sha1'
-require 'ae_easy-core'
+require 'ae_easy/core'
 require 'ae_easy/text/version'
 
 module AeEasy
@@ -34,21 +34,23 @@ module AeEasy
       CGI.unescapeHTML text
     end
 
-    # Strip a value.
+    # Strip a value by trimming spaces, reducing secuential spaces into a
+    #   single space, decode HTML entities and change encoding to UTF-8.
     #
     # @param [String,Object,nil] raw_text Text to strip.
+    # @param [String] orig_encoding Text original encoding.
     #
     # @return [String,nil] `nil` when +raw_text+ is nil, else `String`.
-    def self.strip raw_text
+    def self.strip raw_text, orig_encoding = 'ASCII'
       return nil if raw_text.nil?
       raw_text = raw_text.to_s unless raw_text.is_a? String
       regex = /(\s|\u3000|\u00a0)+/
       good_encoding = (raw_text =~ /\u3000/ || true) rescue false
       unless good_encoding
-        raw_text = raw_text.force_encoding($APP_CONFIG[:encoding]).encode('UTF-8')
+        raw_text = raw_text.force_encoding(orig_encoding).encode('UTF-8', invalid: :replace, undef: :replace)
         regex = /(\s|\u3000|\u00a0|\u00c2\u00a0)+/
       end
-      text = raw_text&.gsub(regex, ' ')&.strip
+      text = raw_text.gsub(regex, ' ').strip
       text.nil? ? nil : decode_html(text)
     end
 
@@ -58,8 +60,9 @@ module AeEasy
     # @param [Hash] data Data hash to save parsed data into.
     # @param [String,Symbol] key Header column key being parsed.
     def self.default_parser cell_element, data, key
-      cell_element&.search('//i').remove
-      row_data[key] = strip cell_element&.text
+      return if cell_element.nil?
+      cell_element.search('//i').remove if cell_element.search('//i').count > 0
+      data[key] = strip cell_element.text
     end
 
     # Parse row data matching a selector using a header map to translate
@@ -74,6 +77,8 @@ module AeEasy
     #   index dictionary.
     # @option opts [Hash{Symbol,String => lambda,proc}] :column_parsers ({})
     #   Custom column parsers for advance data extraction.
+    # @option opts [Boolean] :ignore_text_nodes (true) Ignore text nodes when
+    #   retriving content cells and rows.
     #
     # @yieldparam [Hash{Symbol,String => Object}] data Parsed row data.
     # @yieldparam [Array] row Raw row data.
@@ -87,7 +92,8 @@ module AeEasy
         selector: nil,
         first_row_header: false,
         header_map: {},
-        column_parsers: {}
+        column_parsers: {},
+        ignore_text_nodes: true
       }.merge opts
 
       # Setup config
@@ -96,10 +102,13 @@ module AeEasy
       first = first_row_header = opts[:first_row_header]
       header_map = opts[:header_map]
       column_parsers = opts[:column_parsers]
+      ignore_text_nodes = opts[:ignore_text_nodes]
 
       # Get and parse rows
       html_rows = opts[:html].css(opts[:selector])
       html_rows.each do |row|
+        next if ignore_text_nodes && row.name == 'text'
+
         # First row header validation
         if first && first_row_header
           first = false
@@ -110,7 +119,9 @@ module AeEasy
         row_data = {}
         header_map.each do |key, index|
           # Parse column html with default or custom parser
-          child_element = row.children[index]
+          children = row.children
+          children = children.select{|i|i.name != 'text'} if ignore_text_nodes
+          child_element = children[index]
           column_parsers[key].nil? ?
             default_parser(child_element, row_data, key) :
             column_parsers[key].call(child_element, row_data, key)
@@ -129,12 +140,13 @@ module AeEasy
     #
     # @return [Symbol,String] Translated key.
     def self.translate_label_to_key element, label_map
-      element&.search('//i').remove
-      text = strip element&.text
-      key = label_map.find do |k,v|
+      return nil if element.nil?
+      element.search('//i').remove if element.search('//i').count > 0
+      text = strip element.text
+      key_pair = label_map.find do |k,v|
         v.is_a?(Regexp) ? (text =~ v) : (text == v)
-      end&.first
-      key
+      end
+      key = key_pair.nil? ? nil : key_pair[0]
     end
 
     # Parse header from selector and create a header map to match a column key
@@ -147,6 +159,8 @@ module AeEasy
     #   Key vs. label dictionary.
     # @option opts [Boolean] :first_row_header (false) If true then selector
     #   first matching row will be used as header for parsing.
+    # @option opts [Boolean] :ignore_text_nodes (true) Ignore text nodes when
+    #   retriving header cells and rows.
     #
     # @return [Hash{Symbol,String => Integer},nil] Key vs. column index map.
     def self.parse_header_map opts = {}
@@ -154,11 +168,13 @@ module AeEasy
         html: nil,
         selector: nil,
         column_key_label_map: {},
-        first_row_header: false
+        first_row_header: false,
+        ignore_text_nodes: true
       }.merge opts
 
       # Setup config
       dictionary = opts[:column_key_label_map]
+      ignore_text_nodes = opts[:ignore_text_nodes]
       data = []
       column_map = nil
 
@@ -167,8 +183,12 @@ module AeEasy
       return nil if html_rows.nil?
       html_rows = [html_rows.first] if opts[:first_row_header]
       html_rows.each do |row|
+        next if ignore_text_nodes && row.name == 'text'
+
         column_map = {}
-        row.children.each_with_index do |col, index|
+        children = row.children
+        children = children.select{|i|i.name != 'text'} if ignore_text_nodes
+        children.each_with_index do |col, index|
           # Parse and map column header
           column_key = translate_label_to_key col, dictionary
           next if column_key.nil?
@@ -192,6 +212,8 @@ module AeEasy
     #   first matching row will be used as header for parsing.
     # @option opts [Hash{Symbol,String => lambda,proc}] :column_parsers ({})
     #   Custom column parsers for advance data extraction.
+    # @option opts [Boolean] :ignore_text_nodes (true) Ignore text nodes when
+    #   retriving cells and rows.
     #
     # @yieldparam [Hash{Symbol,String => Object}] data Parsed content row data.
     # @yieldparam [Array] row Raw content row data.
@@ -208,19 +230,22 @@ module AeEasy
         header_key_label_map: {},
         content_selector: nil,
         first_row_header: false,
-        column_parsers: {}
+        column_parsers: {},
+        ignore_text_nodes: true
       }.merge opts
       return nil if opts[:html].nil?
       header_map = self.parse_header_map html: opts[:html],
         selector: opts[:header_selector],
         column_key_label_map: opts[:header_key_label_map],
-        first_row_header: opts[:first_row_header]
+        first_row_header: opts[:first_row_header],
+        ignore_text_nodes: opts[:ignore_text_nodes]
       return nil if header_map.nil?
       data = self.parse_content html: opts[:html],
         selector: opts[:content_selector],
         header_map: header_map,
         first_row_header: opts[:first_row_header],
         column_parsers: opts[:column_parsers],
+        ignore_text_nodes: opts[:ignore_text_nodes],
         &filter
       {header_map: header_map, data: data}
     end
@@ -237,6 +262,8 @@ module AeEasy
     # @option opts [String] :content_selector Content row elements selector.
     # @option opts [Hash{Symbol,String => lambda,proc}] :column_parsers ({})
     #   Custom column parsers for advance data extraction.
+    # @option opts [Boolean] :ignore_text_nodes (true) Ignore text nodes when
+    #   retriving cells and rows.
     #
     # @yieldparam [Hash{Symbol,String => Object}] data Parsed content row data.
     # @yieldparam [Array] row Raw content row data.
@@ -253,7 +280,8 @@ module AeEasy
         header_selector: nil,
         header_key_label_map: {},
         content_selector: nil,
-        column_parsers: {}
+        column_parsers: {},
+        ignore_text_nodes: true
       }.merge opts
       return nil if opts[:html].nil?
 

data/lib/ae_easy/text/version.rb

@@ -1,6 +1,6 @@
 module AeEasy
   module Text
     # Gem version
-    VERSION = "0.0.1"
+    VERSION = "0.0.2"
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: ae_easy-text
 version: !ruby/object:Gem::Version
-  version: 0.0.1
+  version: 0.0.2
 platform: ruby
 authors:
 - Eduardo Rosales
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2019-02-26 00:00:00.000000000 Z
+date: 2019-03-08 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: ae_easy-core
@@ -16,14 +16,14 @@ dependencies:
     requirements:
     - - ">="
       - !ruby/object:Gem::Version
-        version: '0'
+        version: 0.1.2
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
       - !ruby/object:Gem::Version
-        version: '0'
+        version: 0.1.2
 - !ruby/object:Gem::Dependency
   name: bundler
   requirement: !ruby/object:Gem::Requirement