Integrate income_statement_ferc1 table

[cmgosnell]

Overview

Closes #1813
This tbl is weird because it has two dbf tabels flowing into one pudl table.

• bespoke extract step the concatenates the two dbf tables together
• pudl metadata resource and field definitions
• row maps for the dbf row numbers (which is 348 of the line changes!)
• both the dbf and xbrl table need a reshape (dbf table has columns for each utility_type while the xbrl has multiple columns for each income_type
• I enabled align_row_numbers_dbf to take a list of dbf_table_names instead of a single dbf_table_name. This felt a little silly because this many be only one table that ever needs this but it was very easy to implement and felt simpler than having two version of align_row_numbers_dbf
• I had to make an overridden version of source_table_id because the Ferc1AbstractTableTransformer assumes there is only one source table. The new version uses the table name that was added during the extract step. We cooooould do this for all of the DBF tables, but we would need to always add the XBRL table name directly from the extract step. That all felt too.. much for one table. The main quirk here is that source_table_id in the main assign_record_id method takes a source_ferc1 and a df. for all of the other tables the df does nada, but the income statement table needs it. The default method as kwargs so this works okay but feels a little weird.

PR Checklist

Before requesting a review of your pull request, please make sure you've done the
following:

• Merge the most recent version of dev (or the appropriate upstream branch) into
  your branch and resolved any merge conflicts. You may need to do this several
  times over the course of a PR as dev changes frequently.
• Verify that all of the CI checks on your PR are passing. See
  Running Tests with Tox
  for details on how to run the full test suite locally if you need to debug a
  particular failure.
• Ensure that the docstrings for any new modules, classes, functions, or methods are
  descriptive enough for developers and users to understand your code.
• If you expanded data coverage or changed the outputs, ensure that the full
  data validation tests
  pass locally on a fresh DB.
• If you've added new functions or classes, ensure that they have at least basic
  unit tests.
• If you've added new analyses, make sure they include defensive sanity checks that
  will catch unexpected data issues.
• Update the
  release notes
  to reflect your changes. Make sure to reference the PR and any related issues.
• Do your own review of the PR. Add comments highlighting areas where you have
  questions you'd like reviewers to answer, known issues, solutions you're
  unsatisfied with, or other things that deserve special attention from the
  reviewer.

[codecov]

Codecov Report

Base: 85.3% // Head: 85.4% // Increases project coverage by +0.0% 🎉

Coverage data is based on head (cbff8fa) compared to base (c959d4b).
Patch coverage: 89.5% of modified lines in pull request are covered.

Additional details and impacted files

@@          Coverage Diff          @@
##             dev   #2147   +/-   ##
=====================================
  Coverage   85.3%   85.4%           
=====================================
  Files         73      73           
  Lines       8746    8777   +31     
=====================================
+ Hits        7469    7496   +27     
- Misses      1277    1281    +4     

Impacted Files	Coverage Δ
src/pudl/metadata/fields.py	100.0% <ø> (ø)
src/pudl/metadata/resources/ferc1.py	100.0% <ø> (ø)
src/pudl/transform/params/ferc1.py	100.0% <ø> (ø)
src/pudl/transform/ferc1.py	94.7% <88.0%> (-0.5%)	⬇️
src/pudl/extract/ferc1.py	86.0% <100.0%> (+0.2%)	⬆️

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.

[zaneselvans]

I guess we never figured out why you were having weird Numpy failures in the CI. If the builds pass would you bump the max Numpy version back up to <1.25?

[zaneselvans]

Hey these changes look good. Hopefully the numpy thing isn't an issue!