Add Plutus V4 reference script witness flags

[carbolymer]

Context

The issue asks for cardano-cli to support Plutus V4 scripts:

• Ensure Plutus V4 scripts can be deserialised by cardano-cli #1448
  Reading a PlutusScriptV4 text envelope through the --*-script-file options already works, because those readers have been version agnostic since cardano-api 11.7.0.0.
  The missing piece was a --*-plutus-script-v4 flag for each reference script witness kind, matching the existing v2 and v3 flags.
  This PR adds that flag for the spending, minting, withdrawal, certificate, vote and proposal reference script witnesses.

Plutus V4 reference script witnesses have a wire format in Dijkstra transactions, so they work end to end.
A V4 script supplied directly as a witness cannot be exercised yet, because Dijkstra's transaction witness set has no field for V4 scripts and the ledger silently drops such scripts instead of rejecting them:

• Add a field for PlutusV4 scripts to transaction_witness_set CDDL cardano-ledger#5403
  For that reason this PR adds no golden test with a direct V4 witness; it would pin a malformed transaction as expected output.
  The companion cardano-api PR makes cardano-api fail instead of dropping an unsupported inline script, and accepts V4 in Dijkstra in the legacy era check, which --tx-out-reference-script-file needs before it can carry a V4 script:

• Add Plutus V4 support in the Dijkstra era cardano-api#1359

Addresses #1448.
The direct witness part of the issue waits for the ledger change.

How to trust this PR

Run cardano-cli dijkstra transaction build-raw with any of the new --*-plutus-script-v4 flags, for example --spending-plutus-script-v4 alongside --spending-tx-in-reference, and inspect the result with cardano-cli debug transaction view.
The V4 test fixture, v4-always-succeeds.plutus, is the V3 always-succeeds script relabelled as V4.
That is the same approach cardano-api's own generators use, because plutus-ledger-api 1.65 has no real V4 script to draw on.

Two golden tests cover the change.
One exercises all six new flags through dijkstra transaction build-raw and debug transaction view.
The other exercises hash script on a V4 envelope.
The help goldens are regenerated for the new flags.

Pending before review: the golden output files and the regenerated help goldens are not pushed yet, and the build-raw test still carries a placeholder for the V4 script hash.

Checklist

• Commit sequence broadly makes sense and commits have useful messages
• New tests are added if needed and existing tests are updated. See Running tests for more details
• Self-reviewed the diff
• Changelog fragment added in .changes/