Uploaded image for project: 'MariaDB Server'
  1. MariaDB Server
  2. MDEV-10764 PL/SQL parser - Phase 2
  3. MDEV-10573

Oracle style multi-table UPDATE syntax

Details

    • Technical task
    • Status: Open (View Workflow)
    • Major
    • Resolution: Unresolved
    • None
    • None
    • OTHER

    Description

      Implement Oracle-style multi-table UPDATE syntax:

      UPDATE (SELECT ... FROM t1,t2 WHERE ... ) SET ...;

      Example:

      DROP TABLE t1;
      		 
      DROP TABLE t2;
      		 
      CREATE TABLE t1 (id INT PRIMARY KEY, p1 INT);
      		 
      INSERT INTO t1 VALUES (1, 100);
      		 
      INSERT INTO t1 VALUES (2, 100);
      		 
      CREATE TABLE t2 (id INT, p2 INT);
      		 
      INSERT INTO t2 VALUES (1, 10);
      		 
      INSERT INTO t2 VALUES (1, 20);
      		 
      UPDATE (SELECT p1,p2 FROM t1,t2 WHERE t1.id=t2.id AND p1=100) SET p2=p2+1;
      		 
      SELECT * FROM t2;


      	ID	   P2
      		 
      ---------- ----------
      		 
      	 1	   11
      		 
      	 1	   21

      Note, Oracle has some limitations. If I rewrite the above UPDATE statement as:

      UPDATE (SELECT p1,p2 FROM t1,t2 WHERE t1.id=t2.id AND p1=100) SET p1=p1+1;

      it returns an error:

      UPDATE (SELECT p1,p2 FROM t1,t2 WHERE t1.id=t2.id AND p1=100) SET p1=p1+1
      		 
                                                                        *
      		 
      ERROR at line 1:
      		 
      ORA-01779: cannot modify a column which maps to a non key-preserved table

      Attachments

        Activity

          Ascending order - Click to sort in descending order
          bar Alexander Barkov created issue -
          bar Alexander Barkov made changes -
          Field Original Value New Value
          Description Implement Oracle-style multi-table {{UPDATE}} syntax:
          {code:sql}
          UPDATE (SELECT ... FROM t1,t2 WHERE ... ) SET ...;
          {code}

          {code:sql}
          DROP TABLE t1;
          DROP TABLE t2;
          CREATE TABLE t1 (id INT PRIMARY KEY, p1 INT);
          INSERT INTO t1 VALUES (1, 100);
          INSERT INTO t1 VALUES (2, 100);
          CREATE TABLE t2 (id INT, p2 INT);
          INSERT INTO t2 VALUES (1, 10);
          INSERT INTO t2 VALUES (1, 20);
          UPDATE (SELECT p1,p2 FROM t1,t2 WHERE t1.id=t2.id AND p1=10) SET p2=p2+10;
          SELECT * FROM t2;
          {code}
          		Implement Oracle-style multi-table {{UPDATE}} syntax:
          {code:sql}
          UPDATE (SELECT ... FROM t1,t2 WHERE ... ) SET ...;
          {code}

          {code:sql}
          DROP TABLE t1;
          DROP TABLE t2;
          CREATE TABLE t1 (id INT PRIMARY KEY, p1 INT);
          INSERT INTO t1 VALUES (1, 100);
          INSERT INTO t1 VALUES (2, 100);
          CREATE TABLE t2 (id INT, p2 INT);
          INSERT INTO t2 VALUES (1, 10);
          INSERT INTO t2 VALUES (1, 20);
          UPDATE (SELECT p1,p2 FROM t1,t2 WHERE t1.id=t2.id AND p1=100) SET p2=p2+10;
          SELECT * FROM t2;
          {code}
          bar Alexander Barkov made changes -
          Description Implement Oracle-style multi-table {{UPDATE}} syntax:
          {code:sql}
          UPDATE (SELECT ... FROM t1,t2 WHERE ... ) SET ...;
          {code}

          {code:sql}
          DROP TABLE t1;
          DROP TABLE t2;
          CREATE TABLE t1 (id INT PRIMARY KEY, p1 INT);
          INSERT INTO t1 VALUES (1, 100);
          INSERT INTO t1 VALUES (2, 100);
          CREATE TABLE t2 (id INT, p2 INT);
          INSERT INTO t2 VALUES (1, 10);
          INSERT INTO t2 VALUES (1, 20);
          UPDATE (SELECT p1,p2 FROM t1,t2 WHERE t1.id=t2.id AND p1=100) SET p2=p2+10;
          SELECT * FROM t2;
          {code}
          		Implement Oracle-style multi-table {{UPDATE}} syntax:
          {code:sql}
          UPDATE (SELECT ... FROM t1,t2 WHERE ... ) SET ...;
          {code}

          {code:sql}
          DROP TABLE t1;
          DROP TABLE t2;
          CREATE TABLE t1 (id INT PRIMARY KEY, p1 INT);
          INSERT INTO t1 VALUES (1, 100);
          INSERT INTO t1 VALUES (2, 100);
          CREATE TABLE t2 (id INT, p2 INT);
          INSERT INTO t2 VALUES (1, 10);
          INSERT INTO t2 VALUES (1, 20);
          UPDATE (SELECT p1,p2 FROM t1,t2 WHERE t1.id=t2.id AND p1=100) SET p2=p2+1;
          SELECT * FROM t2;
          {code}
          bar Alexander Barkov made changes -
          Description Implement Oracle-style multi-table {{UPDATE}} syntax:
          {code:sql}
          UPDATE (SELECT ... FROM t1,t2 WHERE ... ) SET ...;
          {code}

          {code:sql}
          DROP TABLE t1;
          DROP TABLE t2;
          CREATE TABLE t1 (id INT PRIMARY KEY, p1 INT);
          INSERT INTO t1 VALUES (1, 100);
          INSERT INTO t1 VALUES (2, 100);
          CREATE TABLE t2 (id INT, p2 INT);
          INSERT INTO t2 VALUES (1, 10);
          INSERT INTO t2 VALUES (1, 20);
          UPDATE (SELECT p1,p2 FROM t1,t2 WHERE t1.id=t2.id AND p1=100) SET p2=p2+1;
          SELECT * FROM t2;
          {code}
          		Implement Oracle-style multi-table {{UPDATE}} syntax:
          {code:sql}
          UPDATE (SELECT ... FROM t1,t2 WHERE ... ) SET ...;
          {code}

          {code:sql}
          DROP TABLE t1;
          DROP TABLE t2;
          CREATE TABLE t1 (id INT PRIMARY KEY, p1 INT);
          INSERT INTO t1 VALUES (1, 100);
          INSERT INTO t1 VALUES (2, 100);
          CREATE TABLE t2 (id INT, p2 INT);
          INSERT INTO t2 VALUES (1, 10);
          INSERT INTO t2 VALUES (1, 20);
          UPDATE (SELECT p1,p2 FROM t1,t2 WHERE t1.id=t2.id AND p1=100) SET p2=p2+1;
          SELECT * FROM t2;
          {code}
          {noformat}
          ID P2
          ---------- ----------
          1 11
          1 21
          {noformat}
          bar Alexander Barkov made changes -
          Description Implement Oracle-style multi-table {{UPDATE}} syntax:
          {code:sql}
          UPDATE (SELECT ... FROM t1,t2 WHERE ... ) SET ...;
          {code}

          {code:sql}
          DROP TABLE t1;
          DROP TABLE t2;
          CREATE TABLE t1 (id INT PRIMARY KEY, p1 INT);
          INSERT INTO t1 VALUES (1, 100);
          INSERT INTO t1 VALUES (2, 100);
          CREATE TABLE t2 (id INT, p2 INT);
          INSERT INTO t2 VALUES (1, 10);
          INSERT INTO t2 VALUES (1, 20);
          UPDATE (SELECT p1,p2 FROM t1,t2 WHERE t1.id=t2.id AND p1=100) SET p2=p2+1;
          SELECT * FROM t2;
          {code}
          {noformat}
          ID P2
          ---------- ----------
          1 11
          1 21
          {noformat}
          		Implement Oracle-style multi-table {{UPDATE}} syntax:
          {code:sql}
          UPDATE (SELECT ... FROM t1,t2 WHERE ... ) SET ...;
          {code}

          Example:
          {code:sql}
          DROP TABLE t1;
          DROP TABLE t2;
          CREATE TABLE t1 (id INT PRIMARY KEY, p1 INT);
          INSERT INTO t1 VALUES (1, 100);
          INSERT INTO t1 VALUES (2, 100);
          CREATE TABLE t2 (id INT, p2 INT);
          INSERT INTO t2 VALUES (1, 10);
          INSERT INTO t2 VALUES (1, 20);
          UPDATE (SELECT p1,p2 FROM t1,t2 WHERE t1.id=t2.id AND p1=100) SET p2=p2+1;
          SELECT * FROM t2;
          {code}
          {noformat}
          ID P2
          ---------- ----------
          1 11
          1 21
          {noformat}
          bar Alexander Barkov made changes -
          Description Implement Oracle-style multi-table {{UPDATE}} syntax:
          {code:sql}
          UPDATE (SELECT ... FROM t1,t2 WHERE ... ) SET ...;
          {code}

          Example:
          {code:sql}
          DROP TABLE t1;
          DROP TABLE t2;
          CREATE TABLE t1 (id INT PRIMARY KEY, p1 INT);
          INSERT INTO t1 VALUES (1, 100);
          INSERT INTO t1 VALUES (2, 100);
          CREATE TABLE t2 (id INT, p2 INT);
          INSERT INTO t2 VALUES (1, 10);
          INSERT INTO t2 VALUES (1, 20);
          UPDATE (SELECT p1,p2 FROM t1,t2 WHERE t1.id=t2.id AND p1=100) SET p2=p2+1;
          SELECT * FROM t2;
          {code}
          {noformat}
          ID P2
          ---------- ----------
          1 11
          1 21
          {noformat}
          		Implement Oracle-style multi-table {{UPDATE}} syntax:
          {code:sql}
          UPDATE (SELECT ... FROM t1,t2 WHERE ... ) SET ...;
          {code}

          Example:
          {code:sql}
          DROP TABLE t1;
          DROP TABLE t2;
          CREATE TABLE t1 (id INT PRIMARY KEY, p1 INT);
          INSERT INTO t1 VALUES (1, 100);
          INSERT INTO t1 VALUES (2, 100);
          CREATE TABLE t2 (id INT, p2 INT);
          INSERT INTO t2 VALUES (1, 10);
          INSERT INTO t2 VALUES (1, 20);
          UPDATE (SELECT p1,p2 FROM t1,t2 WHERE t1.id=t2.id AND p1=100) SET p2=p2+1;
          SELECT * FROM t2;
          {code}
          {noformat}
          ID P2
          ---------- ----------
          1 11
          1 21
          {noformat}


          Note, Oracle has some limitations. If I rewrite the above {{UPDATE}} statement as:
          {code:sql}
          UPDATE (SELECT p1,p2 FROM t1,t2 WHERE t1.id=t2.id AND p1=100) SET p1=p1+1;
          {code}
          it returns an error:
          {noformat}
          UPDATE (SELECT p1,p2 FROM t1,t2 WHERE t1.id=t2.id AND p1=100) SET p1=p1+1
                                                                            *
          ERROR at line 1:
          ORA-01779: cannot modify a column which maps to a non key-preserved table
          {noformat}
          monty Michael Widenius made changes -
          Parent MDEV-10142 [ 56873 ]
          Issue Type Task [ 3 ] Technical task [ 7 ]
          monty Michael Widenius added a comment -

          This looks like something that can be done only at the parser level where we convert Oracle syntax to MariaDB multi-table update syntax.

          monty Michael Widenius added a comment - This looks like something that can be done only at the parser level where we convert Oracle syntax to MariaDB multi-table update syntax.
          alvinr Alvin Richards (Inactive) made changes -
          Labels Compatibility
          alvinr Alvin Richards (Inactive) made changes -
          Labels Compatibility
          alvinr Alvin Richards (Inactive) made changes -
          Labels Compatibility
          alvinr Alvin Richards (Inactive) made changes -
          Labels Compatibility
          serg Sergei Golubchik made changes -
          Labels Compatibility
          alvinr Alvin Richards (Inactive) made changes -
          Labels Compatibility Compatibility NRE-307517
          alvinr Alvin Richards (Inactive) made changes -
          NRE Projects NRE-307517
          alvinr Alvin Richards (Inactive) made changes -
          Labels Compatibility NRE-307517 Compatibility
          bar Alexander Barkov made changes -
          Fix Version/s 10.3 [ 22126 ]
          Fix Version/s 10.2 [ 14601 ]
          elenst Elena Stepanova made changes -
          Comment [ Should it have 10.2 as a fix version? ]
          alvinr Alvin Richards (Inactive) made changes -
          Parent MDEV-10142 [ 56873 ] MDEV-10764 [ 57940 ]
          serg Sergei Golubchik made changes -
          Fix Version/s 10.3 [ 22126 ]
          ralf.gebhardt Ralf Gebhardt made changes -
          Support case ID not-26307
          ralf.gebhardt Ralf Gebhardt made changes -
          NRE Projects AC-2610/DEFERRED
          serg Sergei Golubchik made changes -
          Workflow MariaDB v3 [ 76678 ] MariaDB v4 [ 140036 ]

          People

            Unassigned Unassigned
            bar Alexander Barkov
            Votes:
            0 Vote for this issue
            Watchers:
            5 Start watching this issue

            Dates

              Created:
              Updated:

              Git Integration

                Error rendering 'com.xiplink.jira.git.jira_git_plugin:git-issue-webpanel'. Please contact your Jira administrators.