somandal commented on code in PR #8953:
URL: https://github.com/apache/pinot/pull/8953#discussion_r905600098
##########
pinot-segment-spi/src/main/java/org/apache/pinot/segment/spi/index/reader/ForwardIndexReader.java:
##########
@@ -410,9 +421,303 @@ default byte[] getBytes(int docId, T context) {
/**
* MULTI-VALUE COLUMN RAW INDEX APIs
- * TODO: Not supported yet
*/
+ /**
+ * Fills the values
+ * @param docIds Array containing the document ids to read
+ * @param length Number of values to read
+ * @param values Values to fill
+ * @param context Reader context
+ */
+ default void readValuesMV(int[] docIds, int length, int[][] values, T
context) {
+ switch (getValueType()) {
+ case INT:
+ for (int i = 0; i < length; i++) {
+ values[i] = getIntMV(docIds[i], context);
+ }
+ break;
+ case LONG:
+ for (int i = 0; i < length; i++) {
+ long[] longValueBuffer = getLongMV(docIds[i], context);
Review Comment:
I am not sure I fully understand what you mean by reuse the buffer here. Can
you elaborate?
This is my understanding. In this function, we need to create a buffer based
on the `valueType` available. i.e. if the valueType is double, we need to
create a double[] buffer and if it is string, we need a String[] buffer. We
cannot know beforehand which buffer type to create as it depends on
`getValueType()`.
Even if I pass `maxNumValuesPerMVEntry`, I'll have to allocate a typed
buffer here. How is this different from me creating the buffer as part of
`getLongMV()` or one of the other APIs? I will just create it here, pass it to
the `getLongMV()` function, typecast the values in the filled in buffer to the
desired values[][] type and store it in values[][].
Or are you suggesting that we create a buffer depending on the `valueType`
of the caller (e.g. `DataFetcher` knows the `valueType`) within the caller and
pass that pre-allocated buffer of `maxNumValuesPerMVEntry` size down to the
`ForwardIndexReader`? I have 2 questions regarding this, a) will the
`valueType` in the callers like `DataFetcher` match the `valueType` of the
ForwardIndexReader for all purposes? and b) the number of APIs we need to add
to the `ForwardIndexReader` will explode as we'll need one for each buffer type
+ values[][] type combination. Is this acceptable?
I may be misunderstanding your comment, so do let me know what you mean.
--
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: commits-unsubscr...@pinot.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@pinot.apache.org
For additional commands, e-mail: commits-h...@pinot.apache.org
