I have just checked the relevant part of the source: crawler.py The publish_date extraction is currently commented out
# TODO
# article.publish_date = config.publishDateExtractor.extract(doc)
Further examination revealed that if you uncomment the line above, you'll be able to define your custom date extractor. However, there is no default date extractor implemented in Goose. See this method: set_publishdate_extractor
in https://github.com/grangier/python-goose/blob/master/goose/configuration.py