Support GroupConcat sql for aggregating multiple shards in opengauss …

…and doris database(#33797)
apache · Dec 10, 2024 · fdcdf04 · fdcdf04
1 parent f01105d
commit fdcdf04
Show file tree

Hide file tree

Showing 11 changed files with 57 additions and 22 deletions.
diff --git a/RELEASE-NOTES.md b/RELEASE-NOTES.md
@@ -28,7 +28,7 @@
 1. Proxy Native: Change the Base Docker Image of ShardingSphere Proxy Native - [#33263](https://github.com/apache/shardingsphere/issues/33263)
 1. Proxy Native: Support connecting to HiveServer2 with ZooKeeper Service Discovery enabled in GraalVM Native Image - [#33768](https://github.com/apache/shardingsphere/pull/33768)
 1. Proxy Native: Support local transactions of ClickHouse under GraalVM Native Image - [#33801](https://github.com/apache/shardingsphere/pull/33801)
-1. Sharding: Support MYSQL GroupConcat function for aggregating multiple shards - [#33808](https://github.com/apache/shardingsphere/pull/33808)
+1. Sharding: Support GroupConcat function for aggregating multiple shards in Mysql, OpenGauss, Doris - [#33808](https://github.com/apache/shardingsphere/pull/33808)
 1. Proxy Native: Support Seata AT integration under Proxy Native in GraalVM Native Image - [#33889](https://github.com/apache/shardingsphere/pull/33889)
 1. Agent: Simplify the use of Agent's Docker Image - [#33356](https://github.com/apache/shardingsphere/pull/33356)
 

diff --git a/parser/sql/dialect/doris/src/main/antlr4/imports/doris/BaseRule.g4 b/parser/sql/dialect/doris/src/main/antlr4/imports/doris/BaseRule.g4
@@ -965,8 +965,16 @@ udfFunction
     : functionName LP_ (expr? | expr (COMMA_ expr)*) RP_
     ;
 
+separatorName
+    : SEPARATOR string_
+    ;
+
+aggregationExpression
+    : expr (COMMA_ expr)* | ASTERISK_
+    ;
+
 aggregationFunction
-    : aggregationFunctionName LP_ distinct? (expr (COMMA_ expr)* | ASTERISK_)? collateClause? RP_ overClause?
+    : aggregationFunctionName LP_ distinct? aggregationExpression? collateClause? separatorName? RP_ overClause?
     ;
 
 // DORIS ADDED BEGIN

diff --git a/...a/org/apache/shardingsphere/sql/parser/doris/visitor/statement/DorisStatementVisitor.java b/...a/org/apache/shardingsphere/sql/parser/doris/visitor/statement/DorisStatementVisitor.java
@@ -958,14 +958,18 @@ public final ASTNode visitJsonTableFunction(final JsonTableFunctionContext ctx)
 
     private ASTNode createAggregationSegment(final AggregationFunctionContext ctx, final String aggregationType) {
         AggregationType type = AggregationType.valueOf(aggregationType.toUpperCase());
+        String separator = null;
+        if (null != ctx.separatorName()) {
+            separator = new StringLiteralValue(ctx.separatorName().string_().getText()).getValue();
+        }
         if (null != ctx.distinct()) {
             AggregationDistinctProjectionSegment result =
-                    new AggregationDistinctProjectionSegment(ctx.getStart().getStartIndex(), ctx.getStop().getStopIndex(), type, getOriginalText(ctx), getDistinctExpression(ctx));
-            result.getParameters().addAll(getExpressions(ctx.expr()));
+                    new AggregationDistinctProjectionSegment(ctx.getStart().getStartIndex(), ctx.getStop().getStopIndex(), type, getOriginalText(ctx), getDistinctExpression(ctx), separator);
+            result.getParameters().addAll(getExpressions(ctx.aggregationExpression().expr()));
             return result;
         }
-        AggregationProjectionSegment result = new AggregationProjectionSegment(ctx.getStart().getStartIndex(), ctx.getStop().getStopIndex(), type, getOriginalText(ctx));
-        result.getParameters().addAll(getExpressions(ctx.expr()));
+        AggregationProjectionSegment result = new AggregationProjectionSegment(ctx.getStart().getStartIndex(), ctx.getStop().getStopIndex(), type, getOriginalText(ctx), separator);
+        result.getParameters().addAll(getExpressions(ctx.aggregationExpression().expr()));
         return result;
     }
 
@@ -981,11 +985,7 @@ protected Collection<ExpressionSegment> getExpressions(final List<ExprContext> e
     }
 
     private String getDistinctExpression(final AggregationFunctionContext ctx) {
-        StringBuilder result = new StringBuilder();
-        for (int i = 3; i < ctx.getChildCount() - 1; i++) {
-            result.append(ctx.getChild(i).getText());
-        }
-        return result.toString();
+        return ctx.aggregationExpression().getText();
     }
 
     @Override

diff --git a/...a/org/apache/shardingsphere/sql/parser/mysql/visitor/statement/MySQLStatementVisitor.java b/...a/org/apache/shardingsphere/sql/parser/mysql/visitor/statement/MySQLStatementVisitor.java
@@ -22,6 +22,7 @@
 import org.antlr.v4.runtime.ParserRuleContext;
 import org.antlr.v4.runtime.Token;
 import org.antlr.v4.runtime.misc.Interval;
+import org.antlr.v4.runtime.tree.ParseTree;
 import org.antlr.v4.runtime.tree.TerminalNode;
 import org.apache.shardingsphere.sql.parser.api.ASTNode;
 import org.apache.shardingsphere.sql.parser.autogen.MySQLStatementBaseVisitor;
@@ -1041,12 +1042,25 @@ public final ASTNode visitSpecialFunction(final SpecialFunctionContext ctx) {
     public final ASTNode visitGroupConcatFunction(final GroupConcatFunctionContext ctx) {
         calculateParameterCount(ctx.expr());
         FunctionSegment result = new FunctionSegment(ctx.getStart().getStartIndex(), ctx.getStop().getStopIndex(), ctx.GROUP_CONCAT().getText(), getOriginalText(ctx));
-        for (ExprContext each : ctx.expr()) {
+        for (ExprContext each : getTargetRuleContextFromParseTree(ctx, ExprContext.class)) {
             result.getParameters().add((ExpressionSegment) visit(each));
         }
         return result;
     }
 
+    private <T extends ParseTree> Collection<T> getTargetRuleContextFromParseTree(final ParseTree parseTree, final Class<? extends T> clazz) {
+        Collection<T> result = new LinkedList<>();
+        for (int index = 0; index < parseTree.getChildCount(); index++) {
+            ParseTree child = parseTree.getChild(index);
+            if (clazz.isInstance(child)) {
+                result.add(clazz.cast(child));
+            } else {
+                result.addAll(getTargetRuleContextFromParseTree(child, clazz));
+            }
+        }
+        return result;
+    }
+
     @Override
     public final ASTNode visitCastFunction(final CastFunctionContext ctx) {
         FunctionSegment result = new FunctionSegment(ctx.getStart().getStartIndex(), ctx.getStop().getStopIndex(), ctx.CAST().getText(), getOriginalText(ctx));

diff --git a/parser/sql/dialect/opengauss/src/main/antlr4/imports/opengauss/BaseRule.g4 b/parser/sql/dialect/opengauss/src/main/antlr4/imports/opengauss/BaseRule.g4
@@ -954,13 +954,16 @@ typeList
     : typeName (COMMA_ typeName)*
     ;
 
+separatorName
+    : SEPARATOR STRING_
+    ;
+
 funcApplication
     : funcName LP_ RP_
-    | funcName LP_ funcArgList sortClause? RP_
+    | funcName LP_ DISTINCT? funcArgList sortClause? separatorName? RP_
     | funcName LP_ VARIADIC funcArgExpr sortClause? RP_
     | funcName LP_ funcArgList COMMA_ VARIADIC funcArgExpr sortClause? RP_
     | funcName LP_ ALL funcArgList sortClause? RP_
-    | funcName LP_ DISTINCT funcArgList sortClause? RP_
     | funcName LP_ ASTERISK_ RP_
     ;
 

diff --git a/parser/sql/dialect/opengauss/src/main/antlr4/imports/opengauss/Keyword.g4 b/parser/sql/dialect/opengauss/src/main/antlr4/imports/opengauss/Keyword.g4
@@ -664,3 +664,6 @@ MAXVALUE
     : M A X V A L U E
     ;
 
+SEPARATOR
+    : S E P A R A T O R
+    ;
diff --git a/...ache/shardingsphere/sql/parser/opengauss/visitor/statement/OpenGaussStatementVisitor.java b/...ache/shardingsphere/sql/parser/opengauss/visitor/statement/OpenGaussStatementVisitor.java
@@ -449,7 +449,7 @@ public ASTNode visitFuncExpr(final FuncExprContext ctx) {
         Collection<ExpressionSegment> expressionSegments = getExpressionSegments(getTargetRuleContextFromParseTree(ctx, AExprContext.class));
         // TODO replace aggregation segment
         String aggregationType = ctx.funcApplication().funcName().getText();
-        if (AggregationType.isAggregationType(aggregationType)) {
+        if (AggregationType.isAggregationType(aggregationType) && null == ctx.funcApplication().sortClause()) {
             return createAggregationSegment(ctx.funcApplication(), aggregationType, expressionSegments);
         }
         FunctionSegment result = new FunctionSegment(ctx.getStart().getStartIndex(), ctx.getStop().getStopIndex(), ctx.funcApplication().funcName().getText(), getOriginalText(ctx));
@@ -594,13 +594,17 @@ public ASTNode visitBExpr(final BExprContext ctx) {
 
     private ProjectionSegment createAggregationSegment(final FuncApplicationContext ctx, final String aggregationType, final Collection<ExpressionSegment> expressionSegments) {
         AggregationType type = AggregationType.valueOf(aggregationType.toUpperCase());
+        String separator = null;
+        if (null != ctx.separatorName()) {
+            separator = new StringLiteralValue(ctx.separatorName().STRING_().getText()).getValue();
+        }
         if (null == ctx.DISTINCT()) {
-            AggregationProjectionSegment result = new AggregationProjectionSegment(ctx.getStart().getStartIndex(), ctx.getStop().getStopIndex(), type, getOriginalText(ctx));
+            AggregationProjectionSegment result = new AggregationProjectionSegment(ctx.getStart().getStartIndex(), ctx.getStop().getStopIndex(), type, getOriginalText(ctx), separator);
             result.getParameters().addAll(expressionSegments);
             return result;
         }
         AggregationDistinctProjectionSegment result =
-                new AggregationDistinctProjectionSegment(ctx.getStart().getStartIndex(), ctx.getStop().getStopIndex(), type, getOriginalText(ctx), getDistinctExpression(ctx));
+                new AggregationDistinctProjectionSegment(ctx.getStart().getStartIndex(), ctx.getStop().getStopIndex(), type, getOriginalText(ctx), getDistinctExpression(ctx), separator);
         result.getParameters().addAll(expressionSegments);
         return result;
     }

diff --git a/test/e2e/sql/src/test/resources/cases/dql/e2e-dql-select-aggregate.xml b/test/e2e/sql/src/test/resources/cases/dql/e2e-dql-select-aggregate.xml
@@ -145,11 +145,11 @@
         <assertion parameters="abc:String" expected-data-source-name="read_dataset" />
     </test-case>
 
-    <test-case sql="SELECT GROUP_CONCAT(o.remark) as order_id_group_concat FROM t_order o where o.order_id > 1 - 1" db-types="MySQL" scenario-types="db,tbl,dbtbl_with_readwrite_splitting,readwrite_splitting,db_tbl_sql_federation">
+    <test-case sql="SELECT GROUP_CONCAT(o.remark) as order_id_group_concat FROM t_order o where o.order_id > 1 - 1" db-types="MySQL,openGauss" scenario-types="db,tbl,dbtbl_with_readwrite_splitting,readwrite_splitting,db_tbl_sql_federation">
         <assertion expected-data-source-name="read_dataset" />
     </test-case>
 
-    <test-case sql="SELECT GROUP_CONCAT(distinct o.remark SEPARATOR ' ') as order_id_group_concat FROM t_order o where o.order_id > 1 - 1" db-types="MySQL" scenario-types="db,tbl,dbtbl_with_readwrite_splitting,readwrite_splitting">
+    <test-case sql="SELECT GROUP_CONCAT(distinct o.remark SEPARATOR ' ') as order_id_group_concat FROM t_order o where o.order_id > 1 - 1" db-types="MySQL,openGauss" scenario-types="db,tbl,dbtbl_with_readwrite_splitting,readwrite_splitting">
         <assertion expected-data-source-name="read_dataset" />
     </test-case>
 </e2e-test-cases>
diff --git a/test/it/parser/src/main/resources/case/dml/select-special-function.xml b/test/it/parser/src/main/resources/case/dml/select-special-function.xml
@@ -28,6 +28,9 @@
                         <parameter>
                             <column name="status" start-index="20" stop-index="25" />
                         </parameter>
+                        <parameter>
+                            <column name="status" start-index="36" stop-index="41" />
+                        </parameter>
                     </function>
                 </expr>
             </expression-projection>

diff --git a/test/it/parser/src/main/resources/sql/supported/dml/select-aggregate.xml b/test/it/parser/src/main/resources/sql/supported/dml/select-aggregate.xml
@@ -38,6 +38,6 @@
     <sql-case id="select_approx_count"
               value="select owner, approx_count(*) , approx_rank(partition by owner order by approx_count(*) desc) from t group by owner having approx_rank(partition by owner order by approx_count(*) desc) &lt;= 1 order by 1"
               db-types="Oracle"/>
-    <sql-case id="select_group_concat" value="SELECT GROUP_CONCAT(user_id) AS user_id_group_concat FROM t_order" db-types="MySQL"/>
-    <sql-case id="select_group_concat_with_distinct_with_separator" value="SELECT GROUP_CONCAT(distinct user_id SEPARATOR ' ') AS user_id_group_concat FROM t_order" db-types="MySQL"/>
+    <sql-case id="select_group_concat" value="SELECT GROUP_CONCAT(user_id) AS user_id_group_concat FROM t_order" db-types="MySQL,Doris,openGauss"/>
+    <sql-case id="select_group_concat_with_distinct_with_separator" value="SELECT GROUP_CONCAT(distinct user_id SEPARATOR ' ') AS user_id_group_concat FROM t_order" db-types="MySQL,Doris,openGauss"/>
 </sql-cases>
diff --git a/test/it/parser/src/main/resources/sql/supported/dml/select-special-function.xml b/test/it/parser/src/main/resources/sql/supported/dml/select-special-function.xml
@@ -17,7 +17,7 @@
   -->
 
 <sql-cases>
-    <sql-case id="select_group_concat_with_order_by" value="SELECT GROUP_CONCAT(status ORDER BY status) FROM t_order" db-types="MySQL" />
+    <sql-case id="select_group_concat_with_order_by" value="SELECT GROUP_CONCAT(status ORDER BY status) FROM t_order" db-types="MySQL,Doris,openGauss" />
     <sql-case id="select_window_function" value="SELECT order_id, ROW_NUMBER() OVER() FROM t_order" db-types="MySQL" />
     <sql-case id="select_cast_function" value="SELECT CAST('1' AS UNSIGNED)" db-types="MySQL" />
     <sql-case id="select_cast" value="SELECT CAST(c AT TIME ZONE 'UTC' AS DATETIME)" db-types="MySQL" />