[amazon] fixes for DE

[moritzj29]

Hi, there have been some minor changes on Amazon DE webpage, which broke the invoice scraping. Also I fixed handling archived orders for non EN languages.

[Zburatorul]

@moritzj29 could you resolve the conflicts please?

[moritzj29]

Sorry for the delayed response, unfortunately there have been issues again with the link extraction for hidden invoices (DE), which should be fixed now.

Regarding the conflict:

<<<<<<< dev/amazon
                (order_filter,), = self.wait_and_return(
                    lambda: self.find_visible_elements(By.XPATH, '//select[@name="timeFilter"]')
                )
=======
                order_filter, = self.wait_and_locate((By.CSS_SELECTOR, '#time-filter, #orderFilter'))
>>>>>>> master

#97 introduced getting the order filter by a CSS_SELECTOR. In my case, this does not work reliably in headless mode. I myself adjusted the existing XPATH to the updated page. I have no issues with XPATH, also in headless mode. Therefore I would argue for going with this original approach. Please let me know if you think otherwise.

[Zburatorul]

Tested. Works.