Issues with Python Selenium WebElements and .text

Question

I'm writing a text in Selenium that gets the left-most column of a table and verifies that the strings in their cells match a list of dates I have. The code I have looks something like this:

dates = ["20130501", "20130502", "20130506", "20130507", "20130508", "20130509", "20130510", "20130513", "20130514", "20130515"]
mytable = self.driver.find_element(By.ID, "mytable")
datecells = mytable.find_elements(By.CSS_SELECTOR, "tbody td:first-child")
for date, cell in zip(dates, datecells):
    print "'{0}', '{1}'".format(date, cell.text) # For debugging
    #self.assertEqual(date, cell.text)

When the assert is left commented out, I get this printed out as a result:

'20130501', ''
"20130502', ''
'20130506', ''
'20130507', ''
'20130508', ''
'20130509', ''
'20130510', ''
'20130513', ''
'20130514', ''
'20130515', ''

The weird thing is, if I put a breakpoint on the print (Using MyEclipse with PyDev), and look at cell in the variables tab of PyDev Debug before it's output, I can see the proper text, and the code outputs as expected:

'20130501', '20130501'
'20130502', '20130502'
'20130506', '20130506'
'20130507', '20130507'
'20130508', '20130508'
'20130509', '20130509'
'20130510', '20130510'
'20130513', '20130513'
'20130514', '20130514'
'20130515', '20130515'

Is there some weird observer effect quirk of WebElement .text properties where it can only be properly evaluated by this debugger, or is there some condition I should be waiting for in order to get the proper values from the cells without having to step through?

It seems like anything during or after the `datecells` assignment (including stepping into WebDriver's execute function and expanding its result before it's returned), I'm able to break, look at the elements in the debugger, and continue, resulting in the correct values. — TransatlanticFoe, Jun 18 '13 at 13:27

score 4 · Accepted Answer · edited May 23 '17 at 11:44

After testing a little bit with PhantomJS's WebDriver, I was able to figure out this was an issue with Firefox's WebDriver specifically, and found this previously asked question: WebElement getText() is an empty string in Firefox if element is not physically visible on the screen

So after changing the for loop to

for date, cell in zip(dates, datecells):
    self.driver.execute_script("arguments[0].scrollIntoView(true);", cell)
    print "'{0}', '{1}'".format(date, cell.text)
    self.assertEqual(date, cell.text)

in order to scroll each cell into view before getting its text, I got the expected output.

Issues with Python Selenium WebElements and .text

1 Answers1