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

yiguolei pushed a commit to branch branch-4.2
in repository https://gitbox.apache.org/repos/asf/doris.git


The following commit(s) were added to refs/heads/branch-4.2 by this push:
     new 93a21ae81a2 branch-4.2: [improvement](scan) Align scanner split to 
segments for MATCH #67915 (#70775)
93a21ae81a2 is described below

commit 93a21ae81a2b6cce0702f1cd576dc833b014c277
Author: Chenyang Sun <[email protected]>
AuthorDate: Fri Nov 20 17:46:03 2026 +0800

    branch-4.2: [improvement](scan) Align scanner split to segments for MATCH 
#67915 (#70775)
    
    pick from master #67915
    
    Co-authored-by: Claude Opus 5 (1M context) <[email protected]>
---
 be/src/exec/operator/olap_scan_operator.cpp        | 39 +++++++++++++++++-----
 be/src/exec/operator/olap_scan_operator.h          |  3 ++
 .../analyzer/test_custom_analyzer1.groovy          |  8 ++---
 .../analyzer/test_ik_analyzer.groovy               | 16 ++++-----
 .../suites/inverted_index_p0/test_stopwords.groovy | 16 ++++-----
 .../suites/search/test_search_dsl_syntax.groovy    |  6 ++--
 6 files changed, 56 insertions(+), 32 deletions(-)

diff --git a/be/src/exec/operator/olap_scan_operator.cpp 
b/be/src/exec/operator/olap_scan_operator.cpp
index 0043a58fd9a..cfcdf630056 100644
--- a/be/src/exec/operator/olap_scan_operator.cpp
+++ b/be/src/exec/operator/olap_scan_operator.cpp
@@ -449,6 +449,17 @@ static bool contains_expr_node_type(const VExprSPtr& expr, 
TExprNodeType::type n
     });
 }
 
+// Find MATCH recursively; ones nested in AND / OR / NOT count too.
+static bool is_match_expr(const VExprSPtr& expr) {
+    DORIS_CHECK(expr != nullptr);
+    if (expr->node_type() == TExprNodeType::MATCH_PRED ||
+        expr->node_type() == TExprNodeType::SEARCH_EXPR || 
expr->can_push_down_to_index()) {
+        return true;
+    }
+    return std::ranges::any_of(expr->children(),
+                               [](const auto& child) { return 
is_match_expr(child); });
+}
+
 static Status validate_residual_scan_conjuncts(RuntimeState* state,
                                                TPushAggOp::type 
push_down_agg_type,
                                                const VExprContextSPtrs& 
conjuncts) {
@@ -579,6 +590,24 @@ bool OlapScanLocalState::_read_mor_as_dup() {
     return p._olap_scan_node.__isset.read_mor_as_dup && 
p._olap_scan_node.read_mor_as_dup;
 }
 
+// Give each segment a scanner of its own for queries like:
+//   SELECT k1 FROM t ORDER BY l2_distance_approximate(embedding, [1.0, 2.0]) 
LIMIT 2
+//   SELECT k1 FROM t WHERE msg MATCH_PHRASE 'error timeout'
+bool OlapScanLocalState::_use_scan_parallelism_by_per_segment() {
+    if (!state()->query_options().__isset.optimize_index_scan_parallelism ||
+        !state()->query_options().optimize_index_scan_parallelism) {
+        return false;
+    }
+    // TODO: Use optimize_index_scan_parallelism for ann range search in the 
future.
+    // Currently, ann topn is enough
+    if (_ann_topn_runtime != nullptr) {
+        return true;
+    }
+    return config::is_cloud_mode() &&
+           std::ranges::any_of(_common_expr_ctxs_push_down,
+                               [](const auto& ctx) { return 
is_match_expr(ctx->root()); });
+}
+
 Status OlapScanLocalState::_init_scanners(std::list<ScannerSPtr>* scanners) {
     if (_scan_ranges.empty()) {
         _eos = true;
@@ -643,15 +672,7 @@ Status 
OlapScanLocalState::_init_scanners(std::list<ScannerSPtr>* scanners) {
         scanner_builder.set_min_rows_per_scanner(min_rows_per_scanner);
         // BM25 statistics must not depend on how a tablet is split into 
scanners.
         scanner_builder.set_share_tablet_collection_statistics(_score_runtime 
!= nullptr);
-        // If the session variable is set, force one scanner per segment.
-        if (state()->query_options().__isset.optimize_index_scan_parallelism &&
-            state()->query_options().optimize_index_scan_parallelism) {
-            // TODO: Use optimize_index_scan_parallelism for ann range search 
in the future.
-            // Currently, ann topn is enough
-            if (_ann_topn_runtime != nullptr) {
-                scanner_builder.set_scan_parallelism_by_segment(true);
-            }
-        }
+        
scanner_builder.set_scan_parallelism_by_segment(_use_scan_parallelism_by_per_segment());
 
         RETURN_IF_ERROR(scanner_builder.build_scanners(*scanners));
         for (auto& scanner : *scanners) {
diff --git a/be/src/exec/operator/olap_scan_operator.h 
b/be/src/exec/operator/olap_scan_operator.h
index 461e13a4cfa..d7354522ff1 100644
--- a/be/src/exec/operator/olap_scan_operator.h
+++ b/be/src/exec/operator/olap_scan_operator.h
@@ -116,6 +116,9 @@ private:
 
     Status _init_scanners(std::list<ScannerSPtr>* scanners) override;
 
+    // Whether each segment should be scanned by a scanner of its own.
+    bool _use_scan_parallelism_by_per_segment();
+
     Status _build_key_ranges_and_filters();
 
     std::vector<std::unique_ptr<TPaloScanRange>> _scan_ranges;
diff --git 
a/regression-test/suites/inverted_index_p0/analyzer/test_custom_analyzer1.groovy
 
b/regression-test/suites/inverted_index_p0/analyzer/test_custom_analyzer1.groovy
index 83d51ad7831..e6c40c479f4 100644
--- 
a/regression-test/suites/inverted_index_p0/analyzer/test_custom_analyzer1.groovy
+++ 
b/regression-test/suites/inverted_index_p0/analyzer/test_custom_analyzer1.groovy
@@ -107,11 +107,11 @@ suite("test_custom_analyzer1", "p0") {
         sql "sync"
         sql """ set enable_segment_limit_pushdown = true; """
 
-        qt_sql """ select * from test_custom_analyzer1 where ch match 'with'; 
"""
-        qt_sql """ select * from test_custom_analyzer1 where ch match 'the'; 
"""
+        order_qt_sql """ select * from test_custom_analyzer1 where ch match 
'with'; """
+        order_qt_sql """ select * from test_custom_analyzer1 where ch match 
'the'; """
 
-        qt_sql """ select * from test_custom_analyzer2 where ch match 'hm'; """
-        qt_sql """ select * from test_custom_analyzer2 where ch match 'bg'; """
+        order_qt_sql """ select * from test_custom_analyzer2 where ch match 
'hm'; """
+        order_qt_sql """ select * from test_custom_analyzer2 where ch match 
'bg'; """
     } finally {
     }
 }
\ No newline at end of file
diff --git 
a/regression-test/suites/inverted_index_p0/analyzer/test_ik_analyzer.groovy 
b/regression-test/suites/inverted_index_p0/analyzer/test_ik_analyzer.groovy
index e806a339fb0..250d8da7b87 100644
--- a/regression-test/suites/inverted_index_p0/analyzer/test_ik_analyzer.groovy
+++ b/regression-test/suites/inverted_index_p0/analyzer/test_ik_analyzer.groovy
@@ -70,17 +70,17 @@ suite("test_ik_analyzer", "p0") {
 
         // Testing ik smart mode
         println "Testing ik smart mode:"
-        qt_sql """ select * from ${tableNameSmart} where content match_phrase 
'北京'; """
-        qt_sql """ select * from ${tableNameSmart} where content match_phrase 
'计算机科学'; """
-        qt_sql """ select * from ${tableNameSmart} where content match_phrase 
'数据库管理系统'; """
-        qt_sql """ select * from ${tableNameSmart} where content match_phrase 
'中华人民共和国'; """
+        order_qt_sql """ select * from ${tableNameSmart} where content 
match_phrase '北京'; """
+        order_qt_sql """ select * from ${tableNameSmart} where content 
match_phrase '计算机科学'; """
+        order_qt_sql """ select * from ${tableNameSmart} where content 
match_phrase '数据库管理系统'; """
+        order_qt_sql """ select * from ${tableNameSmart} where content 
match_phrase '中华人民共和国'; """
 
         // Testing ik max_word mode
         println "Testing ik max_word mode:"
-        qt_sql """ select * from ${tableNameMaxWord} where content 
match_phrase '北京'; """
-        qt_sql """ select * from ${tableNameMaxWord} where content 
match_phrase '计算机科学'; """
-        qt_sql """ select * from ${tableNameMaxWord} where content 
match_phrase '数据库管理系统'; """
-        qt_sql """ select * from ${tableNameMaxWord} where content 
match_phrase '中华人民共和国'; """
+        order_qt_sql """ select * from ${tableNameMaxWord} where content 
match_phrase '北京'; """
+        order_qt_sql """ select * from ${tableNameMaxWord} where content 
match_phrase '计算机科学'; """
+        order_qt_sql """ select * from ${tableNameMaxWord} where content 
match_phrase '数据库管理系统'; """
+        order_qt_sql """ select * from ${tableNameMaxWord} where content 
match_phrase '中华人民共和国'; """
 
     } finally {
         sql "DROP TABLE IF EXISTS ${tableNameSmart}"
diff --git a/regression-test/suites/inverted_index_p0/test_stopwords.groovy 
b/regression-test/suites/inverted_index_p0/test_stopwords.groovy
index 31e30cf6e48..2ff7dd3f556 100644
--- a/regression-test/suites/inverted_index_p0/test_stopwords.groovy
+++ b/regression-test/suites/inverted_index_p0/test_stopwords.groovy
@@ -44,16 +44,16 @@ suite("test_stopwords", "p0"){
         sql "sync"
         sql """ set enable_segment_limit_pushdown = true """
 
-        qt_sql """ select * from ${indexTbName} where b match 'a'; """
-        qt_sql """ select * from ${indexTbName} where b match 'are'; """
-        qt_sql """ select * from ${indexTbName} where b match 'to'; """
+        order_qt_sql """ select * from ${indexTbName} where b match 'a'; """
+        order_qt_sql """ select * from ${indexTbName} where b match 'are'; """
+        order_qt_sql """ select * from ${indexTbName} where b match 'to'; """
 
-        qt_sql """ select * from ${indexTbName} where c match 'a'; """
-        qt_sql """ select * from ${indexTbName} where c match 'are'; """
-        qt_sql """ select * from ${indexTbName} where c match 'to'; """
+        order_qt_sql """ select * from ${indexTbName} where c match 'a'; """
+        order_qt_sql """ select * from ${indexTbName} where c match 'are'; """
+        order_qt_sql """ select * from ${indexTbName} where c match 'to'; """
 
-        qt_sql """ select * from ${indexTbName} where b match_phrase 'like a 
box'; """
-        qt_sql """ select * from ${indexTbName} where c match_phrase 'like a 
box'; """
+        order_qt_sql """ select * from ${indexTbName} where b match_phrase 
'like a box'; """
+        order_qt_sql """ select * from ${indexTbName} where c match_phrase 
'like a box'; """
 
     } finally {
         //try_sql("DROP TABLE IF EXISTS ${testTable}")
diff --git a/regression-test/suites/search/test_search_dsl_syntax.groovy 
b/regression-test/suites/search/test_search_dsl_syntax.groovy
index 34484e80993..2e8b84b3fbb 100644
--- a/regression-test/suites/search/test_search_dsl_syntax.groovy
+++ b/regression-test/suites/search/test_search_dsl_syntax.groovy
@@ -92,16 +92,16 @@ suite("test_search_dsl_syntax", "p0") {
     qt_sql "SELECT /*+SET_VAR(enable_segment_limit_pushdown=true) */ id, title 
FROM ${tableName} WHERE search('title:Machine') ORDER BY id"
     
     // DSL Syntax Test 2: Phrase queries with quotes
-    qt_sql "SELECT /*+SET_VAR(enable_segment_limit_pushdown=true) */ id, title 
FROM ${tableName} WHERE search('title:\"Machine Learning\"')"
+    qt_sql "SELECT /*+SET_VAR(enable_segment_limit_pushdown=true) */ id, title 
FROM ${tableName} WHERE search('title:\"Machine Learning\"') ORDER BY id"
     
     // DSL Syntax Test 3: Wildcard queries
     qt_sql "SELECT /*+SET_VAR(enable_segment_limit_pushdown=true) */ id, title 
FROM ${tableName} WHERE search('title:*Learning') ORDER BY id"
     
     // DSL Syntax Test 4: Prefix queries
-    qt_sql "SELECT /*+SET_VAR(enable_segment_limit_pushdown=true) */ id, title 
FROM ${tableName} WHERE search('title:Data*')"
+    qt_sql "SELECT /*+SET_VAR(enable_segment_limit_pushdown=true) */ id, title 
FROM ${tableName} WHERE search('title:Data*') ORDER BY id"
     
     // DSL Syntax Test 5: Regular expression queries
-    qt_sql "SELECT /*+SET_VAR(enable_segment_limit_pushdown=true) */ id, title 
FROM ${tableName} WHERE search('title:/.*[Dd]ata.*/')"
+    qt_sql "SELECT /*+SET_VAR(enable_segment_limit_pushdown=true) */ id, title 
FROM ${tableName} WHERE search('title:/.*[Dd]ata.*/') ORDER BY id"
     
     // DSL Syntax Test 6: Boolean AND operator
     qt_sql "SELECT /*+SET_VAR(enable_segment_limit_pushdown=true) */ id, title 
FROM ${tableName} WHERE search('category:Technology AND tags:AI') ORDER BY id"


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to