table_parser 0.3.0 → 0.4.0

data/README.txt

@@ -12,20 +12,46 @@ Parsing table could be difficult when its structure contains colspan or rowspan.
 
 * sudo gem install table_parser
 
-== DEVELOPERS:
 
-* Francis Chong francis at ignition dot hk
+== USAGE:
+
+Use TableParser::Table to create parsed HTML table.
+
+For example, following code:
+<pre>
+	html = "&lt;html&gt;&lt;body&gt;&lt;table&gt;&lt;tr&gt;&lt;td&gt;A&lt;/td&gt;&lt;td&gt;B&lt;/td&gt;&lt;/tr&gt;\
+      &lt;tr&gt;&lt;td rowspan=\"2\"&gt;1&lt;/td&gt;&lt;td&gt;2&lt;/td&gt;&lt;/tr&gt; \
+      &lt;tr&gt;&lt;td&gt;3&lt;/td&gt;&lt;/tr&gt;&lt;/table&gt;&lt;/body&gt;&lt;/html&gt;"
+    table = TableParser::Table.new(html, "/html/body/table")
+</pre>
+
+Result in following parsed table:
 
-After checking out the source, run:
+<pre>
+Table&lt;TableColumn&lt;name=A, children=[1],[1]&gt;,TableColumn&lt;name=B, children=[2],[3]&gt;&gt;
+</pre>
+Note the first column contains duplicated item, because the first row contains "rowspan" element. If this is not desired, use following syntax to skip duplication:
+<pre>
+	html = "&lt;html&gt;&lt;body&gt;&lt;table&gt;&lt;tr&gt;&lt;td&gt;A&lt;/td&gt;&lt;td&gt;B&lt;/td&gt;&lt;/tr&gt;\
+      &lt;tr&gt;&lt;td rowspan=\"2\"&gt;1&lt;/td&gt;&lt;td&gt;2&lt;/td&gt;&lt;/tr&gt; \
+      &lt;tr&gt;&lt;td&gt;3&lt;/td&gt;&lt;/tr&gt;&lt;/table&gt;&lt;/body&gt;&lt;/html&gt;"
+    table = TableParser::Table.new(html, "/html/body/table", false)
+</pre>
 
-  $ rake newb
+Which result in following parsed table:
 
-This task will install any missing dependencies, run the tests/specs,
-and generate the RDoc.
+<pre>
+Table&lt;TableColumn&lt;name=A, children=[1],[1]&gt;,TableColumn&lt;name=B, children=[2],[3]&gt;&gt;
+</pre>
+
+
+== DEVELOPERS:
+
+* Francis Chong francis at ignition dot hk
 
 == LICENSE:
 
-(The MIT License)
+The MIT License
 
 Copyright (c) 2010 Ignition Soft
 

data/lib/table_parser/parser.rb

@@ -32,7 +32,7 @@ module TableParser
       headers
     end
     
-    def self.extract_nodes(rows, headers)
+    def self.extract_nodes(rows, headers, duplicate_colspan)
       data = rows.collect do |row|
         row.collect do |ele|
           node = TableNode.new(ele)
@@ -44,18 +44,21 @@ module TableParser
         row = data[row_index]
         row.each_index do |col_index|
           col = row[col_index]
-          headers[col_index].children << col
+          headers[col_index].children << col if col.class != EmptyTableNode
           
           if col.colspan > 1
             col.insert(col_index, TableNode.new(col.element, col.rowspan, col.colspan - 1))
           end
 
           if col.rowspan > 1 && data[row_index+1]
-            data[row_index+1].insert(col_index, TableNode.new(col.element, col.rowspan - 1))
+            if duplicate_colspan
+              data[row_index+1].insert(col_index, TableNode.new(col.element, col.rowspan - 1))
+            else
+              data[row_index+1].insert(col_index, EmptyTableNode.new(col.rowspan - 1))
+            end
           end
         end
-      end
-      
+      end      
       data
     end
     

data/lib/table_parser/table.rb

@@ -1,14 +1,14 @@
 module TableParser
   class Table
     attr_reader :nodes, :columns
-    def initialize(input, xpath_to_table="//table[0]")
+    def initialize(input, xpath_to_table="//table[0]", duplicate_colspan=true)
       table = Parser.extract_table(input, xpath_to_table)
       @columns = Parser.extract_column_headers(table)
-      @nodes = Parser.extract_nodes(table, @columns)
+      @nodes = Parser.extract_nodes(table, @columns, duplicate_colspan)
     end
   
     def to_s
-      "Table<#{@headers.collect{|h| h.to_s }.join("\n")}>"
+      "Table<#{@columns.collect{|h| h.to_s }.join(",")}>"
     end
 
     # get column by index

data/lib/table_parser/table_column.rb

@@ -15,7 +15,7 @@ module TableParser
     end
     
     def to_s
-      "[name=#{text}, children=#{@children.collect{|c| c.to_s }.join(",")}]"
+      "TableColumn<name=#{text}, children=#{@children.collect{|c| c.to_s }.join(",")}>"
     end
     
   end

data/lib/table_parser/table_node.rb

@@ -3,8 +3,7 @@ module TableParser
     attr_reader :element, :text, :rowspan, :colspan
     def initialize(element, rowspan=nil, colspan=nil)
       @element = element
-      @text = element.text
-      
+      @text = element.text      
       @colspan = colspan || element["colspan"].to_i rescue 1
       @rowspan = rowspan || element["rowspan"].to_i rescue 1
     end
@@ -13,4 +12,13 @@ module TableParser
       "[#{@text}]"
     end
   end
+  
+  class EmptyTableNode < TableNode
+    def initialize(rowspan=nil, colspan=nil)
+      @element = nil
+      @text = ""
+      @colspan = colspan || element["colspan"].to_i rescue 1
+      @rowspan = rowspan || element["rowspan"].to_i rescue 1
+    end
+  end
 end

data/lib/table_parser.rb

@@ -4,5 +4,5 @@ require 'table_parser/table'
 require 'table_parser/parser'
 
 module TableParser
-  VERSION = '0.3.0'
+  VERSION = '0.4.0'
 end

data/test/test_table_parser.rb

@@ -8,11 +8,24 @@ class TestTableParser < Test::Unit::TestCase
       <tr><td>3</td></tr></table></body></html>", 
       "/html/body/table"
 
+    puts table
     assert_equal(2, table.columns.size, 'header_count should = 2 ')
     assert_equal(2, table[0].size)
     assert_equal(2, table[1].size)
   end
   
+  def test_parse_rowspan_disable_dup
+    table = TableParser::Table.new "<html><body><table><tr><td>A</td><td>B</td></tr>\
+      <tr><td rowspan=\"2\">1</td><td>2</td></tr> \
+      <tr><td>3</td></tr></table></body></html>", 
+      "/html/body/table", false
+    
+    puts table
+    assert_equal(2, table.columns.size, 'header_count should = 2 ')
+    assert_equal(1, table[0].size)
+    assert_equal(2, table[1].size)
+  end
+  
   def test_parse_colspan
     table = TableParser::Table.new "<html><body><table><tr><td>A</td><td colspan=\"2\">B</td></tr>\
       <tr><td rowspan=\"2\">A1</td><td>B1</td><td>C1</td></tr> \

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification 
 name: table_parser
 version: !ruby/object:Gem::Version 
-  version: 0.3.0
+  version: 0.4.0
 platform: ruby
 authors: 
 - Francis Chong