The U.S. Securities and Exchange Commission requires companies and individuals to submit data in SEC specific formats. Usually this means taking a columnar dataset and converting it to a specific XML schema.

In practice, this usually means paying a company for proprietary filing software that is annoying to use, and is not modifiable.

Here it an open source solution: https://github.com/john-friedman/secfiler

Note that this is a side project. My main project is manipulating SEC data. That project required me to figure out how to parse every SEC XML file into constituent tables. Since I had to do that, I might as well do the reverse -- which is how secfiler came about. The other project is https://github.com/john-friedman/datamule-python.