format html table with rowspan or colspan

html_table_rowspan_colspan.py

from bs4 import BeautifulSoup
'''
format table with rowspan or colspan to [[]] two dimension lists with the same number td in tr.
html表格格式化，有合并单元格的表格拆分出来，组成td元素个数一致的规整表格。
'''
def test_rowspan():
    html_data_row = '''<table border="1">    
    <tr>    <td>Montd</td>    <td>Savings</td>    <td>Savings for holiday!</td>  </tr>  
    <tr>    <td>January</td>    <td>$100</td>    <td rowspan="2">$50</td>  </tr>  
    <tr>    <td>February</td>    <td>$80</td>  </tr>    </table>'''
    html_data_col = '''
    <table width="100%" border="1">

  <tr>
    <td>Montd</td>
    <td>Savings</td>
  </tr>
  <tr>
    <td colspan="2">January</td>
  </tr>
  <tr>
    <td colspan="2">February</td>
  </tr>
</table>
    '''
    html_data = html_data_col
    table_data = [[cell.text for cell in row("td")] 
                  for row in BeautifulSoup(html_data)("tr")]
    print table_data
    tab = BeautifulSoup(html_data)
    for row_num, row in enumerate(tab("tr")):
        for td_num, td in enumerate(row("td")):
            if 'colspan' in td.attrs and td.attrs['colspan'].isdigit():
                colspan = int(td.attrs['colspan'])
                for i in range(1, colspan):
                    table_data[row_num].insert(td_num, td.text)
            if 'rowspan' in td.attrs and td.attrs['rowspan'].isdigit():
                rowspan = int(td.attrs['rowspan'])
                for i in range(1, rowspan):
                    table_data[row_num+i].insert(td_num, td.text)
    print table_data