`parse()` & `read_*()` issues

[joeflack4]

Sub-issues

• Read method: It might also be cool to have a .read_csv() method like pandas instead of the parse() function (Suggestion: MappingSetDataFrame: .to_csv() and read_csv() #517).
• Disambiguation: There could also be better disambiguation between parse() and parse_file(), which do different things, but you can't tell by the function names.
• Metadata lost: The parse() function simply reads and returns a pandas.DataFrame, not a sssom.MappingSetDataFrame.
• Docs: read_csv() and read_pandas() are missing from the Python docs.
  • Edit: They are actually deprecated, but I don't know why.

[matentzn]

Partial duplicate: #517

[matentzn]

I am not opposed to redesign MSDF to look more like pandas.DataFrame overall. I will admit though that this is very low down the list of priorities right now.

I don't remember the point of the parse method, I would always use read_* methods..

[joeflack4]

@matentzn Thanks. Those methods are also not listed in the Python API documentation. I basically had to dig through the codebase until I found a function that worked.

I'll update the issue OP & title to be related to the read_* ones as well.

Edit: read_csv() and read_pandas() are actually deprecated, but I don't know why.