Extend rule concurrent_009 for manually broken lines

[benthie]

Closes #1456

[JHertz5]

I suggest that you add an example below, like there are for the other settings, to demonstrate how the new setting behaves.

[benthie]

Hi @JHertz5 , I first wanted to wait for a feedback on the proposed feature before implementing the test cases. Maybe @jeremiah-c-leary could also comment on whether he thinks this feature makes sense implementing. If we agree that it is desired, I am happy to write the tests and update the documentation accordingly.

[jeremiah-c-leary]

Afternoon @benthie and @JHertz5 ,

@benthie, if I understand your feature...you would like conditional indenting on concurrent assignments based on the line length. You mentioned rule concurrent_011 had to be disabled to meet your desired formatting. I'm wondering if that rule should be updated to check for line length.

Maybe @jeremiah-c-leary could also comment on whether he thinks this feature makes sense implementing.

The new feature enhances the existing rule, so I think it is valuable to implement.

I apologize for the late reply, but things have been crazy over here.

Regards,

--Jeremy

[benthie]

Hi @jeremiah-c-leary and @JHertz5,

I extended the documentation and examples and added a test.

You mentioned rule concurrent_011 had to be disabled to meet your desired formatting. I'm wondering if that rule should be updated to check for line length.

Rule concurrent_011 has the option new_line_after_assign which currently can either be yes (default), no or ignore. It is executed in phase 1 and concurrent_009 in phase 5. Thus, manually broken lines are realigned with the line of the assignment operator by default by concurrent_011 and never seen "broken" by concurrent_009. For it to work with the new option, concurrent_011 has to be disabled currently.

@jeremiah-c-leary Did you have any particular approach in mind regarding the line length check in concurrent_011? Would that be the first rule besides length_001 to consider the line length? At least I don't know of any other so far.

My ideas are:

• Linked to Autofix line length violations #616
• Let's introduce a new value for new_line_after_assign, e.g. autofix_length and assume it is used
• Scenario 1: assignment on same line as <= and line length okay --> new_line_after_assign = no
  --> no change
• Scenario 2: assignment on same line as <= and line length exceeded --> new_line_after_assign = yes
  --> break the line
• Scenario 3: assignment on new line and if moved up then line length okay --> new_line_after_assign = no
  --> move up
• Scenario 4: assignment on new line and if moved up then line length exceeded --> new_line_after_assign = yes
  --> no change
• In Scenario 2 and 4 the line length might still be exceed if the assignment is on a new line, which then will be caught by length_001