high_level: Move code into a directory and splitting out into files.

[programmer290399]

Pain Point

It is currently not possible to import some high-level functionalities in other core parts of dffml like model.py due to formation of circular import chains.

Which blocks implementation of archive support for #662

References

• model: Rename directory property to location #1155 (comment)
• Weekly Sync 2021-07-13@45:46

Proposed Solution

Split the high_level code into a directory and splitting out into files such as high_level/dataflow.py (similar to cli)

A proposed split tree is given below :

high_level
├── __init__.py
├── high_level.py
│   └── _records_to_sources
├── dataflow.py
│   └── run
├── source.py
│   ├── save
│   └── load
└── ml.py
    ├── train
    ├── accuracy
    └── predict

Tree structure made with tree.nathanfriend.io

[programmer290399]

ping @pdxjohnny

[johnandersen777]

Let's get that helper function out of there. Otherwise looks great!

(echo -e '"""\nHelper functions for dealing with internal data types\n"""\n\n'&& grep -A 39 def\ _records -- dffml/high_level.py | sed -e 's/def _r/def r/') | tee dffml/util/internal.py