PANA-5681: Integrate HeatmapIdentifierRegistry into Session Replay for heatmaps [4 of 4]

[gonzalezreal]

What does this PR do?

Wires ViewIdentityResolver into Session Replay to populate permanentId on wireframes. This enables correlating wireframes with RUM action events for heatmap visualization.

Key changes:

• Adds public ViewIdentityProvider interface for Session Replay's view identity resolution
• Adapts internal ViewIdentityResolver (from RUM feature context) via ViewIdentityResolverAdapter
• Adds viewIdentityProvider to MappingContext for use by all wireframe mappers
• Populates permanentId in all wireframe types (Shape, Text, Image, Placeholder, Webview)
• Calls onWindowRefreshed() on each draw pass to index the current view hierarchy

Motivation

Support mobile heatmaps. This is the final piece connecting Session Replay wireframes to RUM actions.

Additional Notes

• Depends on PANA-5681: Integrate HeatmapIdentifierRegistry into RUM for heatmaps [3 of 4] #3205
• Ported from PANA-5681: Heatmaps #3164. Also updates MappingContextForgeryFactory in session-replay-compose and session-replay-material test suites.

Review checklist (to be filled by reviewers)

• Feature or bugfix MUST have appropriate tests (unit, integration, e2e)
• Make sure you discussed the feature or bugfix with the maintaining team in an Issue
• Make sure each commit and the PR mention the Issue number (cf the CONTRIBUTING doc)

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 39a5a2b5af

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[codecov-commenter]

Codecov Report

❌ Patch coverage is 91.45729% with 17 lines in your changes missing coverage. Please review.
✅ Project coverage is 72.19%. Comparing base (ac886da) to head (eeca598).

Files with missing lines	Patch %	Lines
...lay/internal/recorder/HeatmapIdentifierResolver.kt	91.57%	6 Missing and 1 partial ⚠️
...essionreplay/internal/recorder/SnapshotProducer.kt	92.68%	0 Missing and 3 partials ⚠️
...onreplay/internal/LazyHeatmapIdentifierRegistry.kt	90.91%	2 Missing ⚠️
...internal/recorder/DefaultOnDrawListenerProducer.kt	50.00%	2 Missing ⚠️
...nreplay/internal/recorder/SessionReplayRecorder.kt	85.71%	1 Missing and 1 partial ⚠️
.../sessionreplay/internal/processor/NodeFlattener.kt	50.00%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@                 Coverage Diff                  @@
##           feature/heatmaps    #3206      +/-   ##
====================================================
+ Coverage             72.04%   72.19%   +0.14%     
====================================================
  Files                   967      970       +3     
  Lines                 35499    35666     +167     
  Branches               5887     5938      +51     
====================================================
+ Hits                  25575    25747     +172     
+ Misses                 8304     8303       -1     
+ Partials               1620     1616       -4     

Files with missing lines	Coverage Δ
...dog/android/internal/heatmaps/HeatmapIdentifier.kt	77.78% <100.00%> (ø)
.../sessionreplay/internal/DefaultRecorderProvider.kt	94.37% <100.00%> (+0.04%)	⬆️
...replay/internal/SessionReplayRumContextProvider.kt	92.31% <100.00%> (+0.64%)	⬆️
...onreplay/internal/processor/HeatmapWireframeExt.kt	100.00% <100.00%> (ø)
...ssionreplay/internal/processor/MutationResolver.kt	89.45% <100.00%> (+0.04%)	⬆️
...og/android/sessionreplay/internal/recorder/Node.kt	100.00% <100.00%> (ø)
...nternal/recorder/listener/WindowsOnDrawListener.kt	93.18% <100.00%> (+0.50%)	⬆️
...onreplay/internal/utils/SessionReplayRumContext.kt	90.91% <100.00%> (+0.91%)	⬆️
.../sessionreplay/internal/processor/NodeFlattener.kt	94.44% <50.00%> (-5.56%)	⬇️
...onreplay/internal/LazyHeatmapIdentifierRegistry.kt	90.91% <90.91%> (ø)
... and 4 more

... and 35 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 03802d0335

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[0xnm]

I briefly went through the production files and left some comments, I didn't cover tests.

Given the size of this PR we need to have more several reviewers here.

[0xnm]

Do we actually need to support this case? Cannot we print a warning that RUM is not yet initialized during Session Replay start?

By looking at https://docs.datadoghq.com/session_replay/mobile/setup_and_configuration/?platform=android it is not clear from code snippets that RUM should be initialized first, though (it is only mentioned that RUM SDK setup should be completed).

[jonathanmos]

The problem is that without this mechanism, the heatmaps registry is resolved once at SessionReplay.enable() time. If RUM hasn't registered yet then the registry is permanently null and heatmaps are silently disabled for the entire session. Printing a warning means heatmaps are still broken, and the user has no way to recover without restarting the app.

[kikoveiga]

Would it make sense to add permanentId to the _common-wireframe-schema.json file? So we don't need to do these mappings.

[jonathanmos]

It's possible to put permanentId on the common schema but the reason we didn't do this is that it breaks the previous ordering of arguments for the wireframe constructors instead of permanentId being the last argument, and it also wouldn't remove the need for these mappings. The root issue is that .copy() only exists on data classes, not on the Wireframe sealed class, so to stamp permanentId onto a wireframe subtype we have to downcast via when. To avoid that we'd need to touch the schema generator to emit a withPermanentId() method on each subtype that delegates to its own copy().

[hamorillo]

I went through the production code and haven't found any blockers. That said, the PR is quite big (so it's not easy to review), and it's my first time reviewing SR code, so I'm not confident at all, as I spend most of the time trying to understand everything.

I'll try to come back later with fresh eyes.