[doris] 02/03: [Fix](BinaryPrefixPage) stop to read values when current pos reached the end of the page in `BinaryPrefixPageDecoder::next_batch` (#23855)

[kxiao]

This is an automated email from the ASF dual-hosted git repository.

kxiao pushed a commit to branch branch-2.0
in repository https://gitbox.apache.org/repos/asf/doris.git

commit 2fa15cce135f4a3f03e365a2b5646987210b2777
Author: bobhan1 <bh2444151...@outlook.com>
AuthorDate: Wed Sep 6 16:34:38 2023 +0800

    [Fix](BinaryPrefixPage) stop to read values when current pos reached the 
end of the page in `BinaryPrefixPageDecoder::next_batch` (#23855)
---
 be/src/olap/rowset/segment_v2/binary_prefix_page.cpp | 9 +++++++--
 be/test/olap/primary_key_index_test.cpp              | 2 +-
 2 files changed, 8 insertions(+), 3 deletions(-)

diff --git a/be/src/olap/rowset/segment_v2/binary_prefix_page.cpp 
b/be/src/olap/rowset/segment_v2/binary_prefix_page.cpp
index e3e893a0dd..ab9056def1 100644
--- a/be/src/olap/rowset/segment_v2/binary_prefix_page.cpp
+++ b/be/src/olap/rowset/segment_v2/binary_prefix_page.cpp
@@ -23,6 +23,7 @@
 #include <algorithm>
 #include <vector>
 
+#include "common/status.h"
 #include "gutil/port.h"
 #include "gutil/strings/substitute.h"
 #include "util/coding.h"
@@ -114,12 +115,13 @@ const uint8_t* 
BinaryPrefixPageDecoder::_decode_value_lengths(const uint8_t* ptr
 
 Status BinaryPrefixPageDecoder::_read_next_value() {
     if (_cur_pos >= _num_values) {
-        return Status::NotFound("no more value to read");
+        return Status::EndOfFile("no more value to read");
     }
     uint32_t shared_len;
     uint32_t non_shared_len;
     auto data_ptr = _decode_value_lengths(_next_ptr, &shared_len, 
&non_shared_len);
     if (data_ptr == nullptr) {
+        DCHECK(false) << "[BinaryPrefixPageDecoder::_read_next_value] 
corruption!";
         return Status::Corruption("Failed to decode value at position {}", 
_cur_pos);
     }
     _current_value.resize(shared_len);
@@ -215,8 +217,11 @@ Status BinaryPrefixPageDecoder::next_batch(size_t* n, 
vectorized::MutableColumnP
     // read and copy values
     for (size_t i = 0; i < max_fetch; ++i) {
         dst->insert_data((char*)(_current_value.data()), 
_current_value.size());
-        _read_next_value();
         _cur_pos++;
+        // reach the end of the page, should not read the next value
+        if (_cur_pos < _num_values) {
+            RETURN_IF_ERROR(_read_next_value());
+        }
     }
 
     *n = max_fetch;
diff --git a/be/test/olap/primary_key_index_test.cpp 
b/be/test/olap/primary_key_index_test.cpp
index 66b33a607d..37189f22ec 100644
--- a/be/test/olap/primary_key_index_test.cpp
+++ b/be/test/olap/primary_key_index_test.cpp
@@ -128,7 +128,7 @@ TEST_F(PrimaryKeyIndexTest, builder) {
         EXPECT_FALSE(exists);
         auto status = index_iterator->seek_at_or_after(&slice, &exact_match);
         EXPECT_FALSE(exact_match);
-        EXPECT_TRUE(status.is<NOT_FOUND>());
+        EXPECT_TRUE(status.is<ErrorCode::END_OF_FILE>());
     }
 
     // read all key


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