This is an automated email from the ASF dual-hosted git repository.
kakachen pushed a commit to branch orc
in repository https://gitbox.apache.org/repos/asf/doris-thirdparty.git
The following commit(s) were added to refs/heads/orc by this push:
new 89060c7a5e4 [Enhancement] Add ReadRowCount metric.
89060c7a5e4 is described below
commit 89060c7a5e4c6d53a4e4ddc6cccd90ee4b4afeca
Author: Qi Chen <[email protected]>
AuthorDate: Tue Jul 22 20:06:50 2025 +0800
[Enhancement] Add ReadRowCount metric.
---
c++/include/orc/Reader.hh | 1 +
c++/src/Reader.cc | 1 +
c++/src/sargs/SargsApplier.cc | 16 +++++++++++++---
3 files changed, 15 insertions(+), 3 deletions(-)
diff --git a/c++/include/orc/Reader.hh b/c++/include/orc/Reader.hh
index e3abb143b4a..96a431faae8 100644
--- a/c++/include/orc/Reader.hh
+++ b/c++/include/orc/Reader.hh
@@ -60,6 +60,7 @@ namespace orc {
std::atomic<uint64_t> IOBlockingLatencyUs{0};
std::atomic<uint64_t> SelectedRowGroupCount{0};
std::atomic<uint64_t> EvaluatedRowGroupCount{0};
+ std::atomic<uint64_t> ReadRowCount{0};
};
ReaderMetrics* getDefaultReaderMetrics();
diff --git a/c++/src/Reader.cc b/c++/src/Reader.cc
index 13e1b82d4e3..89bdd77bc6e 100644
--- a/c++/src/Reader.cc
+++ b/c++/src/Reader.cc
@@ -1249,6 +1249,7 @@ namespace orc {
continue;
}
} else {
+ contents->readerMetrics->ReadRowCount.fetch_add(rowsInCurrentStripe);
if (filter) {
// read row group statistics and bloom filters of current stripe
loadStripeIndex();
diff --git a/c++/src/sargs/SargsApplier.cc b/c++/src/sargs/SargsApplier.cc
index 2cc3a7cf44b..7bce1a61aa8 100644
--- a/c++/src/sargs/SargsApplier.cc
+++ b/c++/src/sargs/SargsApplier.cc
@@ -122,12 +122,22 @@ namespace orc {
} while (rowGroup != 0);
// update stats
- uint64_t selectedRGs = std::accumulate(
- mNextSkippedRows.cbegin(), mNextSkippedRows.cend(), 0UL,
- [](uint64_t initVal, uint64_t rg) { return rg > 0 ? initVal + 1 :
initVal; });
+ uint64_t selectedRGs = 0;
+ uint64_t selectedReadRows = 0;
+
+ for (size_t i = 0; i < mNextSkippedRows.size(); ++i) {
+ if (mNextSkippedRows[i] > 0) {
+ selectedRGs++;
+ uint64_t rowGroupStartRow = i * mRowIndexStride;
+ uint64_t rowGroupEndRow = std::min((i + 1) * mRowIndexStride,
rowsInStripe);
+ uint64_t rowsInCurrentRG = rowGroupEndRow - rowGroupStartRow;
+ selectedReadRows += rowsInCurrentRG;
+ }
+ }
if (mMetrics != nullptr) {
mMetrics->SelectedRowGroupCount.fetch_add(selectedRGs);
mMetrics->EvaluatedRowGroupCount.fetch_add(groupsInStripe);
+ mMetrics->ReadRowCount.fetch_add(selectedReadRows);
}
return mHasSelected;
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
