aglinxinyuan opened a new pull request, #7277:
URL: https://github.com/apache/texera/pull/7277

   ### What changes were proposed in this PR?
   
   `LakeFSStorageClient` carried four members with no production caller. Each 
was verified dead by a repo-wide grep before deleting:
   
   | Removed | Callers before | Why it is safe |
   | --- | --- | --- |
   | `stagingApi` | none at all | Private field, unreferenced even inside the 
class. `StagingApi` arrived through the wildcard `io.lakefs.clients.sdk._`, so 
no import change is needed. |
   | `removeFileFromRepo` | none | Every other grep hit is 
`GitVersionControlLocalFileStorage.removeFileFromRepo(Path, Path)` — a 
different method, still live. |
   | `retrieveFileContent` | none | Byte-for-byte duplicate of 
`getFileFromRepo`, which stays live (`DatasetFileDocument`, `DatasetResource`). 
|
   | `withCreateVersion` | 1 test | Only `DatasetResourceSpec`, passing an 
empty `operations` block. |
   
   That one call site moves to the live equivalent. Behaviour is identical: the 
block it ran was empty, and the old method hardcoded the same `branchName = 
"main"`.
   
   ```diff
   -    val commit = LakeFSStorageClient.withCreateVersion(repoName, "commit 
all files") {}
   +    val commit = LakeFSStorageClient.createCommit(repoName, "main", "commit 
all files")
   ```
   
   `writeFileToRepo` is deliberately **kept**. It has no production caller 
either, but `DatasetResourceSpec`, `StagedFileCleanupJobSpec` and 
`LakeFSStorageClientMtimeSpec` all use it to stage a file, and with 
`stagingApi` gone it is the only public staging entry point left.
   
   Also fixes the `getFilePresignedUrl` scaladoc, which was copied from 
`retrieveFileContent`: it described retrieving file content and documented no 
return value, on a method that returns a presigned URL `String`.
   
   ### Any related issues, documentation, discussions?
   
   Closes #7276
   
   Touches the same class as #7273, but not the same files — that PR only names 
these members in a "left uncovered on purpose" comment, so there is no 
conflict. Whichever of the two lands second should drop that comment's second 
bullet, since the members it points at will no longer exist.
   
   ### How was this PR tested?
   
   Deletion only — no behaviour change, no assertion edits, so the gate is that 
nothing still refers to these members and the one moved call site still 
compiles.
   
   ```
   sbt "WorkflowCore/Test/compile" "FileService/Test/compile"
   ```
   
   ```
   sbt "WorkflowCore/scalafmtCheck" "WorkflowCore/Test/scalafmtCheck" 
"FileService/Test/scalafmtCheck" "WorkflowCore/scalafix --check" 
"WorkflowCore/Test/scalafix --check" "FileService/Test/scalafix --check"
   ```
   
   Both green. The Compile config is included because the deleted members live 
in main sources, which `WorkflowCore/Test/*` alone would not lint.
   
   The LakeFS-backed suites that call `writeFileToRepo` (`DatasetResourceSpec`, 
`StagedFileCleanupJobSpec`) stand up postgres / MinIO / LakeFS via 
testcontainers and need a Docker daemon, which was not available on this 
machine; CI runs them.
   
   ### Was this PR authored or co-authored using generative AI tooling?
   
   Generated-by: Claude Code (Opus 5)
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to