Signal and variable assignments are not aligned

[imd1]

Environment
VSG 3.1.0, Centos 7.6

Describe the bug
When a process uses signals and variables, their assignments statements are not vertically aligned

Expected behavior
I would expect them to be aligned

Screenshots
Here's an example of VSG fixed code illustrating the problem:

  begin

    if (rising_edge(s_refclk_p)) then
      if (tx_reset_done = '0' or s_adc_aresetn = '0') then
        sysrefcount := 0;
        s_tx0_sysref <= '0';
      else
        if (sysrefcount = 63) then
          if (DAC_SUBCLASS /= 0) then
            s_tx0_sysref <= '1';
          end if;
          sysrefcount := 0;
        else
          s_tx0_sysref <= '0';
          sysrefcount := sysrefcount + 1;

[imd1]

@jeremiah-c-leary : there are separate rules for signals and variables. What's the best way forward? Create a new rule or update an existing rule?

[imd1]

@jeremiah-c-leary: what do you think about my question above?

[jeremiah-c-leary]

Hey @imd1,

That is a good question. We have moved most of the alignment rules "up" the hierarchy. For example, architecture_026 so we could align :'s across multiple types of declarations. I would think it would apply here also.

So we could depricate sequential_005 and variable_assignment_005 and create an architecture rule. Maybe a procedure and function rule also?

Would we always want to align the <= and the :=?

[imd1]

Hey @imd1,

That is a good question. We have moved most of the alignment rules "up" the hierarchy. For example, architecture_026 so we could align :'s across multiple types of declarations. I would think it would apply here also.

So we could depricate sequential_005 and variable_assignment_005 and create an architecture rule. Maybe a procedure and function rule also?

Would we always want to align the <= and the :=?

My gut feeling is yes we would

[imd1]

I am planning to have a go at this ticket when I next get time...are you OK with that?

[jeremiah-c-leary]

I am planning to have a go at this ticket when I next get time...are you OK with that?

Sure, I would take a look at architecture_029. It might be exactly what you would need.

[imd1]

Having seen your solution to #571, I wonder whether it would be more consistent to update sequential_005 in the same way as subprogram_body_400 and deprecate variable_assignment_005?

[jeremiah-c-leary]

Having seen your solution to #571, I wonder whether it would be more consistent to update sequential_005 in the same way as subprogram_body_400 and deprecate variable_assignment_005?

I would deprecate both sequential_005 and variable_assignment_005 and create a process rule. This is the reasoning:

1. sequential rules should apply within signal assignments
2. variable_assignment rules should apply within variable assignments
3. rules that apply across more than one type of "rule" should exist in the rule category that contains what is being operated on
   a) For example, subprogram_body contains both signal and variable assignments, and needs to operate on both of them.
4. subprogram_body_400 already covers functions and procedures

This is something that has evolved as VSG was worked on. There are still some rules that should be moved according to the above statements.

In fact, sequential rules should be renamed to signal_assignment.

You could probably copy subprogram_body_400 and make it a process rule and change oStart and oEnd.

[jeremiah-c-leary]

Hey @imd1,

Not sure if you were able to get to this, but I pushed an update to this to the issue-562 branch.

When you get a chance could you give it a try and let me know what you think?

Thanks,

--Jeremy

[imd1]

Hey @imd1,

Not sure if you were able to get to this, but I pushed an update to this to the issue-562 branch.

When you get a chance could you give it a try and let me know what you think?

Thanks,

--Jeremy

Effectively reviewed pulled request and it looks good from the unit tests - no time to actually try it out, so go for it when you want to

[jeremiah-c-leary]

This has been merged to master and will be part of the 3.4.0 release.