[PR #2037/65293909 backport][stable-2] New Filter plugin from_csv

[patchback]

This is a backport of PR #2037 as merged into main (6529390).

SUMMARY

This filter adapts the csv parsing features from the read_csv module to use directly with ansible text/variables

Fixes: #1881

ISSUE TYPE

• Feature Pull Request

COMPONENT NAME

plugins/filter/from_csv.py
plugins/modules/read_csv.py
plugins/module_utils/csv.py

ADDITIONAL INFORMATION

Feature set and core implementation is pulled directly from read_csv and adapted to the filter structure.
Common functions have been refactored into module_utils/csv.py which both read_csv and from_csv utilize

Notable Changes:

• Return of read_csv is a tuple of list and dict with populated result switching to dict if key is provided. For the purposes of a filter I have chosen to make the result a list of anonymous dicts which can be keyed using jinja e.g. some_var | community.general.from_csv | groupby('key') | flatten is equivalent to keyed output of read_csv
• Since keys have been removed there is no need for the unique parameter
• Updated instances of ansible.module_utils._text.to_text to ansible.module_utils._text.to_native

Integration tests were added which cover basic functions, but could be updated for better coverage.

Unit tests also added for csv.py.

Wasn't sure how to properly credit the author of read_csv so I updated copyright as if it was an update to the original module

[ansibullbot]

cc @dagwieers
click here for bot help