Enable cloud-migration ReplicateData test (CU 135160 TestReplicateDataProperty)

[aholstrup1]

What & why

Codeunit 135160 "Cloud Migration Property Test" has a single test, TestReplicateDataProperty, which asserts that no table is included in cloud migration unless it is on the approved list — i.e. developers must set ReplicateData = false on tables that should not be migrated from on-prem. That test was on the CI disabled-test skip list, so it never ran, and a large backlog of tables shipping with the wrong ReplicateData property accumulated undetected.

This PR re-enables only that test and introduces a mechanism to temporarily exclude the apps that currently still ship offending tables, so the check can be turned back on now while the debt is burned down per-app. The 10 Test Granules permission-set checks in the same test app stay disabled for now (separate follow-up).

What changed

• src/DisabledTests/Tests-NewObjects-Internal/Tests-NewObjects-Internal.DisabledTest.json — restored with the 10 Test Granules methods still disabled; only the TestReplicateDataProperty entry is removed, re-enabling that single test in CI.
• src/Layers/W1/Tests/NewObjectTests-Internal/CloudMigrationPropertyTest.Codeunit.al — adds an app-id exclusion mechanism (mirrors PR Re-enable Data Classification test (CU 135153 TestDataSensitivities) in CI #9114):
  • GetAppsPendingCloudMigrationProperty — a temporary allow-list of app ids (78 apps) that still ship tables with incorrect ReplicateData.
  • IsTablePendingCloudMigrationProperty — resolves an offending table to its owning app via AllObj."App Package ID" -> NAV App Installed App."App ID" and skips it if the app is grandfathered.
  • Any offending table whose owning app is not on the list still fails the test, so new regressions are caught.
• src/System Application/Test/.../TestMediaCleanup.Table.al and RecordReferenceTest.Table.al — set ReplicateData = false directly. These two System Application test tables are the only offenders outside src/Apps, so they are fixed rather than grandfathered.
• .github/AL-Go-Settings.json — adds src/DisabledTests/* to fullBuildPatterns so edits to disabled-test lists trigger a full CI test run.

Linked work

Fixes #

How I validated this

• I read the full diff and it contains only changes I intended.
• I built the affected app(s) locally with no new analyzer warnings.
• I ran the change in Business Central and confirmed it behaves as expected.
• I added or updated tests for the new behavior, or explained below why none are needed.

What I tested and the outcome

The grandfather app-id list was derived empirically from the actual TestReplicateDataProperty failures across all 22 country builds of a prior CI run (543 distinct offending tables mapped to their owning apps). CI on this PR is the validation that the test now passes with the 78 apps grandfathered and the 2 System App tables fixed.

Risk & compatibility

The exclusion list is a temporary bridge and must burn down to zero as each app sets ReplicateData = false on its offending tables. New tables in apps that are not on the list are still enforced, so this does not create a blind spot for other apps. No product runtime behavior changes.

[github-actions]

Could not find a linked ADO work item. Please link one by using the pattern 'AB#' followed by the relevant work item number. You may use the 'Fixes' keyword to automatically resolve the work item when the pull request is merged. E.g. 'Fixes AB#1234'

[github-actions]

Copilot PR Review

Iteration 6 · Outcome: completed

All sub-skills ran to completion (no-knowledge) against the small AL surface in this PR; the diff is dominated by build/AL-Go settings, disabled-test manifests, and a PowerShell script, none of which fall under the al technologies scope of these leaves.

Knowledge source: https://github.com/microsoft/BCQuality@186d8a131465475c79244d994acb872cd5c0d4bf

Orchestrator pre-filter (2 file(s) excluded)

• layer-disabled (knowledge) : 2 file(s)

_{Findings produced by the AL review agent v1.7.3. Reply 👎 on any inline comment to flag false positives.}

[github-actions]

$\textbf{🟠\ High\ Severity\ —\ Error\ Handling} \quad \color{gray}{\texttt{\small Iteration\ 4}}$

Multiple test methods in this codeunit now hand-roll error accumulation with a DotNet ArrayList (Errors.Add(...) inside a loop, then Error(String.Join(...)) at the end) instead of using the platform's ErrorBehavior::Collect feature.

The referenced guidance calls this exact shape an anti-pattern: it reimplements the collection feature, loses each failure's ErrorInfo structure, and skips telemetry classification. The change is functionally reasonable (report every missing permission/table instead of stopping at the first), but the same outcome is achievable by marking a per-item check procedure ErrorBehavior::Collect, invoking it via Codeunit.Run per item, and aggregating with HasCollectedErrors()/GetCollectedErrors(true) as the good sample shows.

Knowledge:

• microsoft/knowledge/error-handling/collect-validation-errors-with-errorbehavior.md

Line mapping was unavailable, so this was posted as an issue comment.

_{👍 useful · ❤️ especially valuable · 👎 wrong - reply with why}