calling `response.text` in any download middleware loaded after scrapy_crawlera middleware results in `*** AttributeError: Response content isn't text`