Use the Master Detail Browser to browse data in parent tables and drill-down to data in related child tables. This is useful when you need to analyze or edit data. For example, suppose you need to view a list of all Nick's Flix stores and the inventory of comedy movies at each store. Using the Master Detail Browser, you specify the table that contains the list of stores as the parent table and specify the table that contains comedy movies as the child table to quickly locate comedies and analyze the inventory.
To browse tables using the Master Detail Browser
Select Tools | Master Detail Browser (ALT+T+A).
Drag one or more tables from the Object Explorer, Object Search, or Project Manager window to the Diagram pane at the top of the Master Detail Browser. Press CTRL+click to select multiple tables at once.
Notes:
A line between tables indicates any existing relationships between the selected tables.
At least one parent table (table without any join arrows pointing to it) must be added to the window or an error message displays when returning results.
Join Columns.
Select the checkbox beside each column you want to include.
Note: Any columns included in a join are automatically included in the results.
Click (F9) to retrieve data for the parent table initially. To view data in a child table, click + to the left of each row.
or
Click to retrieve data in the parent and child tables. If there is more than one child table, a tab displays for each child table.
From the data grid, you can:
Edit the data and commit your changes to the database.
Right-click the data grid and select Send To | option to add the data to a report, chart, or pivot grid.
Tips:
Click to align tables and snap them to the grid.
Select the parent table from the drop-down list in the toolbar to view data for a parent table if there are multiple parent tables.
Click X on the table you want to remove in the Diagram pane to remove a table and any conditions (e.g., join, where clause) associated with the table.
The Query Builder enables you to create a query without writing or editing SQL statements. Even if you are familiar with SQL, the graphical interface makes it easier to create relationships and visualize the query.
Click here to view a video of this feature.
Note: This procedure does not cover all of the possible steps of visually building a query. It only covers the steps required to build a query for the scenario.
Scenario Your company wants to improve its global sales, starting with countries where it has a customer base but generally low sales. You need to create a query that identifies countries with less than $500,000 in sales for each of the last four fiscal years. Note: This scenario uses the SH (Sales History) sample schema that comes with Oracle 10g so you can follow the procedure. |
To visually build a query based on this scenario
Drag the following tables from the Object Explorer to the Diagram pane: TIMES, SALES, CUSTOMERS, and COUNTRIES.
Tips:
Table |
Column |
---|---|
COUNTRIES | COUNTRY_NAME |
TIMES | FISCAL_YEAR |
SALES | AMOUNT_SOLD |
Tips:
Select the Aggregate Function field in the AMOUNT_SOLD column and select Sum. This sums all of the sales per fiscal year per country.
Note: If you select a group by clause or aggregate function for a column, the rest of the columns in the query must also have a group by clause or aggregate function.
Select 2001 in the second Constant field and click OK.
Note: The SH schema does not have current data. The 1998-2001 years provide the best data for the purposes of this scenario.
Click .
The query results display in the Results tab. See View Result Sets for more information.
To save the query, right-click the Query Builder tab and select Save File.
You can save a query and any results sets in a Toad editor file (.tef). This file format is useful for building scripts that have large result sets that you do not want to continue executing or for saving results when you have not finished building a script.
See Tutorial: Build a Subquery for a tutorial on creating a subquery.
See Add Subqueries for more information on how to add a subquery.
Use the Cross-Connection Query Builder or the Cross-Connection Editor to create a query that combines data from multiple, dissimilar databases. For example, you can create a query that joins managers from a SQL Server database to their employees in an Oracle database. (This is also often called a heterogeneous or cross-database query.) Toad uses an enhanced query engine to implement advanced execution processes that are specially designed to improve the performance of cross-connection queries.
Before creating a cross-connection query using an Oracle, Teradata or MySQL ODBC connection, please see the special considerations listed at the end of this topic.
Note: This feature is available in the Toad Data Point Professional Edition only.
To build a cross-connection query visually
Select Tools | Build Query | Cross-Connection Query Builder.
Note: A blue banner along the right side of the Diagram pane indicates that you are in Cross-Connection Query mode.
Complete the query in the same manner as you would a non-cross-connection query, but note the following exceptions.
Object Explorer Feature | In Cross-Connection Query Mode... |
---|---|
Options |
Connection Treelist is the only view option available. In this option, connections, databases, tables, and views are displayed in a treelist in the Object Explorer window. |
Filters |
Filters are especially useful in cross-connection queries to reduce the number of objects displayed in the treelist in the Object Explorer window. |
Note: Review the "Considerations and Limitations of Cross-Connection Queries" section below.
To compose a cross-connection query in the editor
Select Tools | Edit | Cross-Connection SQL Editor.
Note: A blue banner along the right side of the Editor pane indicates that you are in Cross-Connection Query mode.
Compose the SQL statement in the same manner as you would a non-cross-connection query, but note the following exceptions.
SQL Editor Task | In Cross-Connection Query Mode... |
---|---|
Naming tables and views |
To reference a table or view, you must use a cross-connection fully-qualified name. The format for a cross-connection fully-qualified name depends on the database. The following is an example using the SQL Server format:
Note: For the database connection string, use the connection string displayed in the Navigation Manager window for each connection (or in the Name field in Connection Properties). Tip: Copy and paste a SQL statement from a single connection editor to a cross-connection editor and Toad automatically changes the syntax of the statement to include the cross-connection fully-qualified names. |
Using alias names | You must use alias names for tables and columns in a cross-connection query. |
Using code completion |
In Cross-Connection Query mode, the scope of available options in code completion expands to include the databases, tables, and views for all the open database connections. Note: For a list of SQL functions to use when composing a cross-connection SQL statement, see Cross-Connection Query Functions. Tip: To expand/collapse a database node in the code completion list, use CTRL+left arrow and CTRL+right arrow. |
Tips:
Review the following considerations and limitations before creating a cross-connection query.
Consideration/Limitation |
Description |
---|---|
Performance concerns |
A cross-connection query may take longer to execute than a query against a single database because result sets from each database are retrieved and combined to generate a single result set. In the Professional edition of Toad Data Point, the enhanced execution engine utilizes several optimizing techniques to improve the performance of cross-connection queries. |
ODBC driver support |
An ODBC driver can be selected for a single connection through each connection's Properties dialog or globally through Tools | Options | Database. |
Using an Oracle connection in a cross-connection query |
To use an Oracle connection in a cross-connection query, you must install an Oracle ODBC driver. Note: The Oracle Full Install client includes an ODBC driver. However, if you are using the Oracle Instant client (which may not include an ODBC driver), you may need to install an Oracle ODBC driver. |
Oracle Direct Connection | You cannot use an Oracle Direct Connection in a cross-connection query. |
Using a Teradata connection in a cross-connection query | To use a Teradata connection in a cross-connection query, you must install a Teradata ODBC driver. |
Using a MySQL ODBC connection in a cross-connection query | To use a MySQL ODBC connection in a cross-connection query, you must specify a database for the connection. |
Result sets |
You cannot edit result sets. |
Oracle Long Data Type | Cross-connection query does not support Oracle long data types or Oracle LOBs. |
Optimize Cross-Connection Queries
© ALL RIGHTS RESERVED. Términos de uso Privacidad Cookie Preference Center