Reading from bytes or custom reader, not only from file path

[justanotheranonymoususer]

Thanks for the library. I wanted to try it, but I only have the document in memory in my use case. If such feature can be implemented, that would be great.

[KaiserY]

Just cut off a 0.1.18 version and added a new_from_reader function, it takes a reader which is Read + Seek just like underlying zip new function. And read_docx example also updated.