从BR​​标签提取文本

Question

I have been able to extract text using Selenium before, however I'am having trouble with just extracting the numbers between < BR > tags. Here is a sample of the html code.

<DIV class="pagebodydiv">
    <TABLE  CLASS="datadisplaytable" SUMMARY="This table will display needed information." WIDTH="100%">
<TR>
<TD CLASS="nttitle" scope="colgroup" >Working Title</A></TD>
</TR>
<TR>
<TD CLASS="ntdefault">
 Further information on subject
<BR>
    3.000
<BR>
    2.000  
<BR>
<BR>
<BR>
<BR>
<BR>
More information
<BR>
<BR>
</TABLE>

So far I have tried using:

WebElement creditinfo = driver.findElement(By.xpath("//div[@class='pagebodydiv']/text()[preceding-sibling::br]

and Elements numInfo = doc.select("br");

However, I keep running into a NoSuchElementException error, an InvalidSelectorException error, or it just doesn't return anything. Any ideas on how I can get the information?

Answer 1

You actually can select the text nodes between <BR> tags. In HTML (not XHTML) they act as self-closing tags (like <br/> ). Based on that behaviour, you could select all text nodes that have a <BR> tag before and after it using:

//TABLE[@CLASS='datadisplaytable']/TR/TD[@CLASS="ntdefault"]
/text()[preceding-sibling::node()[1][self::BR] 
        and following-sibling::node()[1][self::BR]]

That would select also the blank lines and the character text which is not a number.

You can get rid of the empty space nodes adding a [normalize-space(.) != ''] to the end of the expression (which will now only return three nodes). And you can select which node you want using a positional predicate at the end of the expression ( [1] to select the first node.

The expression below selects the text node containing the value 2.000 :

//TABLE[@CLASS='datadisplaytable']/TR/TD[@CLASS="ntdefault"]
/text()[preceding-sibling::node()[1][self::BR] 
        and following-sibling::node()[1][self::BR]][normalize-space(.) != ''][2]

Note: I'm assuming your source actually has tag names in uppercase, since in XPath <TD> is not the same as <td> . I'm not sure how tolerant Selenium is about this when parsing HTML.

Answer 2

It may help :

  WebElement table =driver.findElement(By.xpath("//table[@class='datadisplaytable']"));
   WebElement tbody=table.findElement(By.tagName("tbody"));
   List<WebElement> rows=tbody.findElements(By.tagName("tr"));
   System.out.println("Row  size:"+rows.size());
   ArrayList<String> list=new ArrayList<>();

   for(int i=0;i<rows.size();i++)
   {
     WebElement column = tbody.findElement(By.xpath("//table[@class='datadisplaytable']/tbody/tr[2]/td"));
     if(column.getText().trim().contains("."))
     {
        System.out.println("text : "+column.getText().trim());
        list.add(column.getText().trim());
     }

   }

Answer 3

I believe that BR are not considered enclosing tags, and so you will not be able to extract the "enclosing text". You will probably have to extract the text enclosed in your TD CLASS="ntdefault" , where all the BR s are going to get translated into newlines. You will then have to perform string manipulation to extract only the parts you are interested in.