rangareddy commented on issue #17342: URL: https://github.com/apache/hudi/issues/17342#issuecomment-5351235385
This issue was reviewed as part of the JIRA-migrated backlog triage (HUDI-8750). **Findings: not done. The flag is still threaded through the whole file-index stack.** Current call sites of `includeLogFiles` on `master`: - `hudi-spark-datasource/hudi-spark-common/src/main/scala/org/apache/hudi/HoodieFileIndex.scala:94` (constructor parameter, defaults to `false`), `:386` (`f.withLogFiles(includeLogFiles)`), `:460` (`if (includeLogFiles) allBaseFilesAndLogFiles else allBaseFiles`) - `hudi-spark-datasource/hudi-spark-common/src/main/scala/org/apache/hudi/HoodieIncrementalFileIndex.scala:40`, `:43`, `:50`, `:63` - `hudi-spark-datasource/hudi-spark-common/src/main/scala/org/apache/hudi/HoodieHadoopFsRelationFactory.scala:273` (`includeLogFiles = true`) - `hudi-spark-datasource/hudi-spark-common/src/main/scala/org/apache/hudi/ExpressionIndexSupport.scala:94`, `:585`, `:592` - and at the model level, `hudi-common/src/main/java/org/apache/hudi/common/model/FileSlice.java:76` and `:110` (`withLogFiles`) So a boolean is still being plumbed through four classes and a model type to decide something that, under the file-group-reader flow, arguably should not be a caller's choice at all. Like the other HUDI-9107 sub-tasks in this batch (#17327, #17341, #17344), the shape of the cleanup depends on how far the FG-reader migration goes, so it is worth sequencing these together rather than picking them off individually. Keeping this open. -- 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]
