优化生成hive sql的逻辑 #5853

alibaba · May 4, 2024 · a8441f4 · a8441f4
1 parent 9003085
commit a8441f4
Show file tree

Hide file tree

Showing 2 changed files with 105 additions and 16 deletions.
diff --git a/core/src/main/java/com/alibaba/druid/sql/visitor/SQLASTOutputVisitor.java b/core/src/main/java/com/alibaba/druid/sql/visitor/SQLASTOutputVisitor.java
@@ -10962,7 +10962,20 @@ protected void printCreateTable(HiveCreateTableStatement x, boolean printSelect)
             printAndAccept(clusteredBy, ",");
             print(')');
         }
-
+        List<SQLSelectOrderByItem> sortedBy = x.getSortedBy();
+        if (sortedBy.size() > 0) {
+            println();
+            print0(ucase ? "SORTED BY (" : "sorted by (");
+            printAndAccept(sortedBy, ", ");
+            print(')');
+        }
+        int buckets = x.getBuckets();
+        if (buckets > 0) {
+            println();
+            print0(ucase ? "INTO " : "into ");
+            print(buckets);
+            print0(ucase ? " BUCKETS" : " buckets");
+        }
         List<SQLExpr> skewedBy = x.getSkewedBy();
         if (skewedBy.size() > 0) {
             println();
@@ -10991,21 +11004,6 @@ protected void printCreateTable(HiveCreateTableStatement x, boolean printSelect)
         Map<String, SQLObject> serdeProperties = x.getSerdeProperties();
         printSerdeProperties(serdeProperties);
 
-        List<SQLSelectOrderByItem> sortedBy = x.getSortedBy();
-        if (sortedBy.size() > 0) {
-            println();
-            print0(ucase ? "SORTED BY (" : "sorted by (");
-            printAndAccept(sortedBy, ", ");
-            print(')');
-        }
-
-        int buckets = x.getBuckets();
-        if (buckets > 0) {
-            println();
-            print0(ucase ? "INTO " : "into ");
-            print(buckets);
-            print0(ucase ? " BUCKETS" : " buckets");
-        }
 
         SQLExprTableSource like = x.getLike();
         if (like != null) {

diff --git a/core/src/test/java/com/alibaba/druid/bvt/sql/hive/issues/Issue5853.java b/core/src/test/java/com/alibaba/druid/bvt/sql/hive/issues/Issue5853.java
@@ -0,0 +1,91 @@
+package com.alibaba.druid.bvt.sql.hive.issues;
+
+import java.util.List;
+
+import com.alibaba.druid.DbType;
+import com.alibaba.druid.sql.ast.SQLStatement;
+import com.alibaba.druid.sql.parser.SQLParserUtils;
+import com.alibaba.druid.sql.parser.SQLStatementParser;
+
+import org.junit.Test;
+
+import static org.junit.Assert.assertEquals;
+
+/**
+ * Hive解析CREATE TABLE的问题
+ *
+ * @see <a href="https://cwiki.apache.org/confluence/display/Hive/LanguageManual+DDL">...</a>
+ */
+public class Issue5853 {
+
+    @Test
+    public void test_parse_create_0() {
+        for (String sql : new String[]{
+            "CREATE TABLE page_view (\n"
+                + "\tviewTime INT,\n"
+                + "\tuserid BIGINT,\n"
+                + "\tpage_url STRING,\n"
+                + "\treferrer_url STRING,\n"
+                + "\tip STRING COMMENT 'IP Address of the User'\n"
+                + ")\n"
+                + "COMMENT 'This is the page view table'\n"
+                + "PARTITIONED BY (\n"
+                + "\tdt STRING,\n"
+                + "\tcountry STRING\n"
+                + ")\n"
+                + "CLUSTERED BY (userid)\n"
+                + "SORTED BY (viewTime)\n"
+                + "INTO 32 BUCKETS\n"
+                + "ROW FORMAT DELIMITED\n"
+                + "\tFIELDS TERMINATED BY '\\001'\n"
+                + "\tCOLLECTION ITEMS TERMINATED BY '\\002'\n"
+                + "\tMAP KEYS TERMINATED BY '\\003'\n"
+                + "STORED AS SEQUENCEFILE;",
+        }) {
+            System.out.println("原始的sql===" + sql);
+            SQLStatementParser parser1 = SQLParserUtils.createSQLStatementParser(sql, DbType.hive);
+            List<SQLStatement> statementList1 = parser1.parseStatementList();
+            String sqleNew = statementList1.get(0).toString();
+            System.out.println("生成的sql===" + sqleNew);
+            assertEquals(sql, sqleNew);
+            SQLStatementParser parser2 = SQLParserUtils.createSQLStatementParser(sqleNew, DbType.hive);
+            List<SQLStatement> statementList2 = parser2.parseStatementList();
+            String sqleNew2 = statementList2.get(0).toString();
+            System.out.println("再次解析生成的sql===" + sqleNew2);
+            assertEquals(sqleNew, sqleNew2);
+        }
+
+    }
+
+    @Test
+    public void test_parse_create_1() {
+        for (String sql : new String[]{
+            "CREATE TABLE db.route(\n"
+                + "od_id string COMMENT 'OD',\n"
+                + "data_dt string COMMENT 'data date')\n"
+                + "CLUSTERED BY (\n"
+                + "od_id)\n"
+                + "INTO 8 BUCKETS\n"
+                + "ROW FORMAT SERDE\n"
+                + "'org.apache.hadoop.hive.ql.io.orc.OrcSerde'\n"
+                + "STORED AS INPUTFORMAT\n"
+                + "'org.apache.hadoop.hive.ql.io.orc.OrcInputFormat'\n"
+                + "OUTPUTFORMAT\n"
+                + "'org.apache.hadoop.hive.ql.io.orc.OrcOutputFormat';",
+
+        }) {
+            System.out.println("原始的sql===" + sql);
+            SQLStatementParser parser1 = SQLParserUtils.createSQLStatementParser(sql, DbType.hive);
+            List<SQLStatement> statementList1 = parser1.parseStatementList();
+            String sqleNew = statementList1.get(0).toString();
+            System.out.println("生成的sql===" + sqleNew);
+            SQLStatementParser parser2 = SQLParserUtils.createSQLStatementParser(sqleNew, DbType.hive);
+            List<SQLStatement> statementList2 = parser2.parseStatementList();
+            String sqleNew2 = statementList2.get(0).toString();
+            System.out.println("再次解析生成的sql===" + sqleNew2);
+            assertEquals(sqleNew, sqleNew2);
+        }
+
+    }
+
+}