Java codegen: clean output directory before generating to prevent orphan accumulation#1623
Merged
Merged
Conversation
…rphan accumulation ## Summary Add a clean step to `java/scripts/codegen/java.ts` that removes the entire generated output directory before writing new files. This ensures that when upstream schema types are renamed or removed, the previously-generated `.java` files are automatically deleted rather than silently accumulating as orphans. ## Changes - **`java/scripts/codegen/java.ts`** — At the start of `main()`, recursively delete `src/generated/java/com/github/copilot/generated/` and recreate the empty directory before invoking any generators. - **`java/src/generated/java/`** — 204 orphan files removed (types no longer emitted by the current schema). These accumulated over multiple `@github/copilot` releases as types were renamed or removed upstream. - **`GeneratedRpcRecordsCoverageTest.java`** — Removed test methods that exercised deleted generated types (`SessionWorkspaceCreateFileParams`, `SessionModeGetResult`, `SessionModeSetResult`, `DiscoveredMcpServerSource`, and 16 others). - **`GeneratedRpcApiCoverageTest.java`** — Removed test method for the deleted `SessionAgentDeselectResult` type. ## Why this works going forward The CI workflow (`java-codegen-check.yml`) runs codegen and then checks `git status --porcelain`. Because codegen now cleans before writing, any future orphans will appear as deletions in the diff and the check will fail — preventing orphan accumulation without any additional workflow changes. Fixes #1619
edburns
changed the title
Contributor
Cross-SDK Consistency Review ✅All 207 changed files are in Why the clean-before-generate pattern is Java-specificThe orphan accumulation problem exists only in the Java codegen because Java conventionally writes one file per class. The other SDK codegens all write to a fixed set of monolithic files (always overwritten, so they cannot accumulate orphans):
The fix (clean the output directory before generating) is correctly scoped to Java. No equivalent change is needed in any other SDK.
|
Collaborator
Author
|
I'm very sorry about this mistake, this PR is ready for review @stephentoub . |
Contributor
There was a problem hiding this comment.
Pull request overview
This PR improves the Java SDK’s code generation workflow by cleaning the Java generated-sources output directory before regenerating from the upstream JSON schemas, preventing stale/orphaned .java files from accumulating when schema types are renamed or removed (fixes #1619).
Changes:
- Add a “clean generated output dir” step at the start of
java/scripts/codegen/java.tsso regeneration reflects the current schema exactly (including deletions). - Remove previously orphaned generated Java types that are no longer emitted by the current
@github/copilotschemas. - Update Java generated-type coverage tests to stop referencing deleted generated types.
Show a summary per file
| File | Description |
|---|---|
| java/scripts/codegen/java.ts | Adds a pre-generation clean step for src/generated/java/com/github/copilot/generated to prevent orphan accumulation. |
| java/src/test/java/com/github/copilot/generated/rpc/GeneratedRpcApiCoverageTest.java | Removes a coverage test that exercised a generated type that is no longer emitted. |
| java/src/generated/java/com/github/copilot/generated/** | Regenerated output: removes stale orphaned generated types no longer present in the schema. |
| java/src/generated/java/com/github/copilot/generated/rpc/** | Regenerated output: removes stale orphaned generated RPC types no longer present in the schema. |
Copilot's findings
- Files reviewed: 1/207 changed files
- Comments generated: 1
Comment on lines
+2062
to
+2066
| // Clean the generated output directory to remove orphaned files from previous runs | ||
| const generatedOutputDir = path.join(REPO_ROOT, "src/generated/java/com/github/copilot/generated"); | ||
| console.log(`🧹 Cleaning output directory: ${generatedOutputDir}`); | ||
| await fs.rm(generatedOutputDir, { recursive: true, force: true }); | ||
| await fs.mkdir(generatedOutputDir, { recursive: true }); |
stephentoub
approved these changes
Jun 10, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Add a clean step to
java/scripts/codegen/java.tsthat removes the entire generated output directory before writing new files. This ensures that when upstream schema types are renamed or removed, the previously-generated.javafiles are automatically deleted rather than silently accumulating as orphans.Changes
java/scripts/codegen/java.ts— At the start ofmain(), recursively deletesrc/generated/java/com/github/copilot/generated/and recreate the empty directory before invoking any generators.java/src/generated/java/— 204 orphan files removed (types no longer emitted by the current schema). These accumulated over multiple@github/copilotreleases as types were renamed or removed upstream.GeneratedRpcRecordsCoverageTest.java— Removed test methods that exercised deleted generated types (SessionWorkspaceCreateFileParams,SessionModeGetResult,SessionModeSetResult,DiscoveredMcpServerSource, and 16 others).GeneratedRpcApiCoverageTest.java— Removed test method for the deletedSessionAgentDeselectResulttype.Why this works going forward
The CI workflow (
java-codegen-check.yml) runs codegen and then checksgit status --porcelain. Because codegen now cleans before writing, any future orphans will appear as deletions in the diff and the check will fail — preventing orphan accumulation without any additional workflow changes.Fixes #1619