jpountz commented on code in PR #14482:
URL: https://github.com/apache/lucene/pull/14482#discussion_r2059102197


##########
lucene/core/src/java/org/apache/lucene/store/Directory.java:
##########
@@ -79,6 +83,31 @@ public abstract class Directory implements Closeable {
    */
   public abstract long fileLength(String name) throws IOException;
 
+  protected void validateIOContext(IOContext context) {
+    Map<Class<? extends IOContext.FileOpenHint>, List<IOContext.FileOpenHint>> 
hintClasses =
+        
context.hints().stream().collect(Collectors.groupingBy(IOContext.FileOpenHint::getClass));
+
+    // there should only be one of FileType, FileData, DataAccess
+    List<IOContext.FileOpenHint> fileTypes =
+        hintClasses.getOrDefault(FileTypeHint.class, List.of());
+    if (fileTypes.size() > 1) {
+      throw new IllegalArgumentException("Multiple file type hints specified: 
" + fileTypes);
+    }
+    List<IOContext.FileOpenHint> fileData = 
hintClasses.getOrDefault(FileDataHint.class, List.of());
+    if (fileData.size() > 1) {
+      throw new IllegalArgumentException("Multiple file data hints specified: 
" + fileData);
+    }
+    List<IOContext.FileOpenHint> dataAccess =
+        hintClasses.getOrDefault(DataAccessHint.class, List.of());
+    if (dataAccess.size() > 1) {
+      throw new IllegalArgumentException("Multiple data access hints 
specified: " + dataAccess);
+    }
+  }
+
+  protected ReadAdvice toReadAdvice(IOContext context) {

Review Comment:
   > After all, you're specifying a specific read advice for a reason - if you 
encode the reason (as hints), rather than the specific behaviour that you want, 
then the directory can choose the best way to do that according to the 
directory implementation
   
   My reading of recent reports is that users want to be able to customize it 
because there isn't a one-size-fits-all. For instance, vectors work better with 
`ReadAdvice.NORMAL` when they can fit in the page cache, and better with 
`ReadAdvice.RANDOM` when they significantly exceed the size of the page cache. 
So I agree that Lucene should come with sensible defaults, but I think that we 
also need a way to diverge from Lucene's defaults?



-- 
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: issues-unsubscr...@lucene.apache.org

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org
For additional commands, e-mail: issues-h...@lucene.apache.org

Reply via email to