------------------ * Added origin, version and release date to scraped information * Changed company and classification to return a list instead of a string
* Handles cases where multiple companies or classifications are involved
0.1.2
------------------ * First release on PyPI. * Allows to scrape
* All figure IDs owned by a particular user * For a particular figure
* The characters represented * The company manufacturing the figure * Its classification * The main image URL
* Images associated to a given figure
* Stores the scraped content as a JSON list in a file. * Stores the figure images.