SOLR-17600 (Part 1/4): Migrate V2 API Payloads from MapSerializable by isaric · Pull Request #4463 · apache/solr

isaric · 2026-05-25T13:20:54Z

This is Part 1 of 4 to deprecate and remove the MapSerializable interface as part of SOLR-17600.

This PR migrates all V2 API payload and handler classes to use MapWriter / Utils.convertToMap.

Note: The work has been split into 4 PRs to make reviewing easier. The subsequent PRs are stacked on top of this one.

Copilot

Pull request overview

Note

Copilot was unable to run its full agentic suite in this review.

Refactors multiple Solr handlers/tests to standardize conversion of Solr/MapWriter-style objects into map-like structures (favoring Utils.convertToMap(...) / SimpleOrderedMap) and updates one internal API adapter to a MapWriter#writeMap(...) implementation.

Changes:

Replaced a number of toMap(new HashMap<>()) call sites with Utils.convertToMap(...) or new SimpleOrderedMap<>(...).
Updated config/response building code paths to use MapWriter / SimpleOrderedMap rather than MapSerializable.
Adjusted BaseHandlerApiSupport parameter adapter to implement writeMap(...) and skip null-valued entries.

Reviewed changes

Copilot reviewed 23 out of 23 changed files in this pull request and generated 8 comments.

Show a summary per file

File	Description
solr/core/src/test/org/apache/solr/handler/designer/TestSchemaDesignerAPI.java	Updates test param map construction to use `SimpleOrderedMap`.
solr/core/src/java/org/apache/solr/handler/export/ExportWriterStream.java	Uses `SimpleOrderedMap` when materializing `MapWriter` values.
solr/core/src/java/org/apache/solr/handler/designer/ManagedSchemaDiff.java	Uses `Utils.convertToMap` for diff map materialization.
solr/core/src/java/org/apache/solr/handler/admin/api/SplitShardAPI.java	Uses `Utils.convertToMap` for v2 payload → v1 params.
solr/core/src/java/org/apache/solr/handler/admin/api/SplitCoreAPI.java	Uses `Utils.convertToMap` for v2 payload → v1 params.
solr/core/src/java/org/apache/solr/handler/admin/api/RequestSyncShardAPI.java	Uses `Utils.convertToMap` for v2 payload → v1 params.
solr/core/src/java/org/apache/solr/handler/admin/api/RequestCoreRecoveryAPI.java	Uses `Utils.convertToMap` for v2 payload → v1 params.
solr/core/src/java/org/apache/solr/handler/admin/api/RequestBufferUpdatesAPI.java	Uses `Utils.convertToMap` for v2 payload → v1 params.
solr/core/src/java/org/apache/solr/handler/admin/api/RequestApplyCoreUpdatesAPI.java	Uses `Utils.convertToMap` for v2 payload → v1 params.
solr/core/src/java/org/apache/solr/handler/admin/api/RejoinLeaderElectionAPI.java	Uses `Utils.convertToMap` for v2 payload → v1 params.
solr/core/src/java/org/apache/solr/handler/admin/api/RebalanceLeadersAPI.java	Uses `Utils.convertToMap` for v2 payload → v1 params.
solr/core/src/java/org/apache/solr/handler/admin/api/PrepareCoreRecoveryAPI.java	Uses `Utils.convertToMap` for v2 payload → v1 params.
solr/core/src/java/org/apache/solr/handler/admin/api/OverseerOperationAPI.java	Builds v1 params using `SimpleOrderedMap` instead of `toMap(HashMap)`.
solr/core/src/java/org/apache/solr/handler/admin/api/MoveReplicaAPI.java	Uses `Utils.convertToMap` for v2 payload → v1 params.
solr/core/src/java/org/apache/solr/handler/admin/api/ModifyCollectionAPI.java	Uses `Utils.convertToMap` for v2 payload → v1 params.
solr/core/src/java/org/apache/solr/handler/admin/api/MigrateDocsAPI.java	Uses `Utils.convertToMap` for v2 payload → v1 params.
solr/core/src/java/org/apache/solr/handler/admin/api/InstallShardData.java	Uses `Utils.convertToMap` before creating `ZkNodeProps`.
solr/core/src/java/org/apache/solr/handler/admin/api/CreateCore.java	Uses `Utils.convertToMap` to derive request body from v1 params.
solr/core/src/java/org/apache/solr/handler/admin/ZookeeperInfoHandler.java	Uses `Utils.convertToMap` for `DocCollection` serialization.
solr/core/src/java/org/apache/solr/handler/admin/IndexSizeEstimator.java	Uses `SimpleOrderedMap` when converting `MapWriter` results.
solr/core/src/java/org/apache/solr/handler/admin/BaseHandlerApiSupport.java	Switches adapter from `toMap(...)` to `writeMap(...)` and filters nulls.
solr/core/src/java/org/apache/solr/handler/SolrConfigHandler.java	Migrates plugin/config map building to `MapWriter`/`SimpleOrderedMap` and adds helper for component lookup.
solr/core/src/java/org/apache/solr/handler/ClusterAPI.java	Builds role command maps using `SimpleOrderedMap`.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

dsmiley · 2026-05-27T18:45:48Z

+        Map<String, Object> map = new SimpleOrderedMap<>(queue);
        entry.setValue(map);
      } else if (value instanceof MapWriterSummaryStatistics stats) {
-        Map<String, Object> map = new LinkedHashMap<>();
-        stats.toMap(map);
+        Map<String, Object> map = new SimpleOrderedMap<>(stats);


SOM is good here; this is a response-structure

Agreed — keeping SOM.

SimpleOrderedMap<T> extends NamedList<T> implements Map<String, T> (SimpleOrderedMap.java:46), so it is a Map. @dsmiley also confirmed SOM is appropriate here as a response structure.

dsmiley · 2026-05-27T19:01:41Z

-                    o instanceof MapSerializable
-                        ? ((MapSerializable) o).toMap(new LinkedHashMap<>())
+                    o instanceof MapWriter
+                        ? new SimpleOrderedMap<>((MapWriter) o)


SOM is okay for response

Keeping SOM here. The resulting pluginInfo map is mutated downstream in getComponentInfo (computeIfAbsent) and to inject _packageinfo_ via put, so it needs to be a real mutable Map.

SimpleOrderedMap implements Map<String, T>, so the type is compatible. The SOM stays here because the resulting map is mutated downstream (computeIfAbsent, put).

isaric · 2026-06-06T12:17:26Z

        pluginInfo = (Map) plugin;
      } else if (plugin instanceof PluginInfo) {
-        pluginInfo = ((PluginInfo) plugin).toMap(new LinkedHashMap<>());
+        pluginInfo = new SimpleOrderedMap<>((PluginInfo) plugin);


SimpleOrderedMap implements Map<String, T>, so the type is compatible. This line was reverted to .toMap(new LinkedHashMap<>()) because PluginInfo does not implement MapWriter until Part 2.

dsmiley

Thanks for resuming this effort!

dsmiley · 2026-05-27T18:45:48Z

+        Map<String, Object> map = new SimpleOrderedMap<>(queue);
        entry.setValue(map);
      } else if (value instanceof MapWriterSummaryStatistics stats) {
-        Map<String, Object> map = new LinkedHashMap<>();
-        stats.toMap(map);
+        Map<String, Object> map = new SimpleOrderedMap<>(stats);


SOM is good here; this is a response-structure

dsmiley · 2026-05-27T19:01:41Z

-                    o instanceof MapSerializable
-                        ? ((MapSerializable) o).toMap(new LinkedHashMap<>())
+                    o instanceof MapWriter
+                        ? new SimpleOrderedMap<>((MapWriter) o)


SOM is okay for response

dsmiley · 2026-06-03T01:59:41Z

Please never force-push once someone has seen/reviewed your PR. It resets the review state. Not only that, but I see it's still one commit so I really don't know what changed.

dsmiley · 2026-06-03T02:22:55Z

The code here seems mostly populating a response. For such cases, I question if you need to convert existing MapWriter things to SimpleOrderedMap (e.g. by using SOM's constructor). Meaning, simply pass the MapWriter as-is. I believe the response writers will all check the value to be of type MapWriter and then do the right thing. I mentioned this point in Part 2 previously.

Applied at line 208 where the value isn't mutated. The remaining SOM conversions (lines 292, 339, 372) all feed code paths that subsequently call computeIfAbsent / put / setValue on the map, so they need a mutable Map and the conversion stays.

Note: 339 and 372 were also broken in the originally pushed code (tried to wrap a MapSerializable in SimpleOrderedMap, which only has a MapWriter constructor — premature change that assumed Part 2's migration). I reverted them to .toMap(new LinkedHashMap<>()) so Part 1 compiles standalone; the SOM-wrap will land naturally in Part 2 once SolrConfig / PluginInfo become MapWriter.

- OverseerOperationAPI: restore HashMap (LinkedHashMap was an unintended change; all sibling *API classes use HashMap) - SolrConfigHandler: drop needless SimpleOrderedMap wrap around ParamSet; response writers handle MapWriter directly - SolrConfigHandler: revert two SimpleOrderedMap conversions that required Part 2's MapWriter migration to compile; restored to toMap(LinkedHashMap) to keep Part 1 independently buildable

isaric · 2026-06-06T12:16:50Z

Sorry about the force-push — that was a mistake. Future updates to all four PRs will land as additive commits so review state is preserved.

dsmiley · 2026-06-06T16:41:14Z

The change is widespread enough across these 4 parts that I think it deserves an "other" type of changelog entry. I guess we'll add that to the last one that actually vanquishes MapSerializable.

dsmiley · 2026-06-06T16:43:09Z

LMK if you've addressed everything and you think this is ready for merge. I don't want to scrutinize this further; I'm glad you looked into the matters I called out.

Copilot AI review requested due to automatic review settings May 25, 2026 13:20

github-actions Bot added the tests label May 25, 2026

This was referenced May 25, 2026

SOLR-17600 (Part 2/4): Migrate Core Configuration classes from MapSerializable #4464

Open

SOLR-17600 (Part 3/4): Migrate SolrJ and Utilities from MapSerializable #4465

Open

SOLR-17600 (Part 4/4): Delete MapSerializable interface #4466

Open

Copilot AI reviewed May 25, 2026

View reviewed changes

isaric force-pushed the SOLR-17600-1-v2-api branch 2 times, most recently from baae128 to 74d671e Compare May 26, 2026 05:55

dsmiley reviewed May 27, 2026

View reviewed changes

SOLR-17600 (Part 1/4): Migrate V2 API Payloads from MapSerializable

00c356e

isaric force-pushed the SOLR-17600-1-v2-api branch from 74d671e to 00c356e Compare June 2, 2026 14:52

dsmiley reviewed Jun 3, 2026

View reviewed changes

Conversation

isaric commented May 25, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

dsmiley left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

dsmiley commented Jun 3, 2026

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

isaric commented Jun 6, 2026

Uh oh!

dsmiley commented Jun 6, 2026

Uh oh!

dsmiley commented Jun 6, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants