Re: TEXT columns should indentify as java.sql.Types.CLOB
От | Toni Helenius |
---|---|
Тема | Re: TEXT columns should indentify as java.sql.Types.CLOB |
Дата | |
Msg-id | E6A9CAA76548CB4EB02D2E3B174DD3B155381F0EAE@ink.sad.syncrontech.com обсуждение исходный текст |
Ответ на | Re: TEXT columns should indentify as java.sql.Types.CLOB (dmp <danap@ttc-cmc.net>) |
Ответы |
Re: TEXT columns should indentify as java.sql.Types.CLOB
|
Список | pgsql-jdbc |
Hi, An definitely a bug: BOOLEAN columns get identified as java.sql.Types.VARCHAR. Not java.sql.Types.BOOLEAN as they should.I end up getting SQL create statements where boolean fields are translated back as VARCHAR(1). -----Original Message----- From: dmp [mailto:danap@ttc-cmc.net] Sent: 16. elokuuta 2010 18:36 To: Toni Helenius; pgsql-jdbc@postgresql.org Subject: Re: [JDBC] TEXT columns should indentify as java.sql.Types.CLOB > > >Hello, > >I'm using Postgres 8.3.11 database and the latest JDBC driver 8.4 Build 701 (JDBC 4) + Java 6. In our databases there areTEXT type columns. However if I make a query to identify these fields in Java, the field DATA TYPE is VARCHAR and thelength is 2147483647. Type name is correct; "TEXT". But as we need database independent code, we are using DATA TYPE asI presume is correct. And I think these TEXT fields should return java.sql.Types.CLOB as DATA TYPE instead of VARCHAR. > >Here is some code: > >Connection fromConn; >DatabaseMetaData metaFrom; >String userFrom; > >fromConn = from.getConnection(); >metaFrom = fromConn.getMetaData(); >userFrom = ((PooledConnection)from).getTableOwner(); > >ResultSet cols = metaFrom.getColumns(null, userFrom, code, null); while >(cols.next()) { > cols.getShort("DATA_TYPE"); > cols.getString("TYPE_NAME"); > } > I'm just not seeing it. Your code example seems to be collecting the information from the database connection not a particulartable. If your application needs to identify column types regardless of different databases then the way I do itis through evaluation of the table columns. Attached file containing the output for the last three or so PostgreSQL JDBCdrivers. As far as TEXT and CLOB types I would prefer then to be identifed independently. danap String sqlStatementString = "SELECT * FROM " + schemaTableName + " LIMIT 1"; ResultSet db_resultSet = sqlStatement.executeQuery(sqlStatementString); DatabaseMetaData dbMetaData = dbConnection.getMetaData(); ResultSetMetaData tableMetaData = db_resultSet.getMetaData(); for (int i = 1; i < tableMetaData.getColumnCount() + 1; i++) { // Collect Information on Column. colNameString = tableMetaData.getColumnName(i); comboBoxNameString = parseColumnNameField(colNameString); columnClass = tableMetaData.getColumnClassName(i); columnType = tableMetaData.getColumnTypeName(i); columnSize = Integer.valueOf(tableMetaData.getColumnDisplaySize(i)); System.out.println(i + " " + colNameString + " " + comboBoxNameString + " " + columnClass + " " + columnType + " " + columnSize);}
В списке pgsql-jdbc по дате отправления: