Date picker submission bug

[zayne-siew]

Issue Description

Date inputs that include the year are automatically converted into a date picker. The HTML element that controls the date picker requires a 'special' submission method using Selenium ActionChains (see src/questions/date.py).

On a local machine, Selenium ActionChains works fine. However, when deployed onto Heroku, the date value selected in the Telegram bot does not reflect onto the submitted form response.

Debugging/Troubleshooting Logs

• day, month and year accurately store the selected day, month and year values from the date parsed. The value submitted using send_keys also gives the expected output:

"{:02d}{:02d}{:04d}".format(day, month, year)  # e.g. day=29, month=9, year=2021 ==> 29092021

• On Heroku, the HTML date picker element is successfully being selected and values are being inputted correctly. Based on local machine testing, this means that both move_to_element_with_offset and send_keys are working (I think), since the form would not be able to submit successfully otherwise.

# On local machine testing, this method will input the key into the year value of the date picker
# and result in an invalid input. Subsequently, the form cannot be submitted.
# e.g. day=29, month=9, year=2021; browser: "dd/mm/2909", where "dd" and "mm" are placeholders in the form HTML.
self._ANSWER_ELEMENTS.click()
self._ANSWER_ELEMENTS.send_keys(key)

# The "special" method used via ActionChains gives the expected output on local machine testing.
# Something similar happens on Heroku deployment, but no idea why the value submitted is incorrect.
# e.g. day=29, month=9, year=2021; local browser: "29/09/2021", heroku: "09/02/2021".
action.move_to_element_with_offset(self._ANSWER_ELEMENTS, 0, 0) \
    .click() \
    .send_keys("{:02d}{:02d}{:04d}".format(day, month, year)) \
    .perform()

Replication Instructions

Use the Telegram bot to submit a response to any Google Form like this one with the abovementioned date picker HTML element, and check the recorded form response.