Fix PsrpOperator constructor validation to use provision checks inste…

[bujjibabukatta]

Summary

Part of the burn-down tracked in #70296.

PsrpOperator.__init__ validated command/powershell/cmdlet/arguments/parameters
using truthiness checks instead of is not None checks. Per the guidance in #70296,
"was this provided" checks are allowed to stay in __init__, but must use is not None
polarity rather than truthiness — so this is a fix in place, not a move to execute().

What was wrong

• command/powershell/cmdlet were collected into a set before being passed to
  exactly_one(). If two of them happened to render to the same string, the set would
  collapse them and validation would incorrectly pass.
• if arguments and ..., if parameters and ..., if cmdlet: treated an explicitly
  provided empty list/dict/string as "not provided", silently skipping validation.

Fix

Replaced truthiness checks with is not None checks, and pass fields to exactly_one()
directly instead of via a set. __init__ is now limited to this provision check plus
plain self.field = field assignments.

Testing

Added three regression tests in test_psrp.py covering: two duplicate field values,
an empty arguments list, and an empty parameters dict — all now correctly raise
ValueError.

Related: #70296

Was generative AI tooling used to co-author this PR?

• Yes - Claude

Generated-by: Claude following the guidelines

[shahar1]

Please note that it's a competitor PR of #70347 (you may keep this open, but I might end up merging the other - depending which provides the best match first)

[eladkal]

Why was this file modified?

[potiuk]

Thanks — your diagnosis is right, and the fix is correct: args = {command, powershell, cmdlet} built a set (collapsing equal values) and exactly_one counted truthiness rather than presence, so the is not None rewrite is the proper correction.

It duplicates #70347 though, which makes the identical change to the same two files and was opened on 23 July, six days before this one. That PR also carries the fix through execute() and removes the operator from validate_operators_init_exemptions.txt, so it's the more complete version — I'm closing this in its favour and have asked there for the one extra behavioural change it contains to be split out or documented.

One unrelated observation while I'm here: this PR also touches scripts/in_container/run_generate_constraints.py, which doesn't appear connected to the PsrpOperator fix. If that was an accidental inclusion it's worth checking your branch; if it's a real fix it deserves its own PR where it'll actually get reviewed rather than riding along unnoticed.

Worth a quick gh pr list --search "<operator>" before starting on one of these — several of the validate_operators_init_exemptions.txt entries have had two people working on them independently this week.

---

Drafted-by: Claude Code (Opus 5); reviewed by @potiuk before posting