Previous Topic Next Topic

Creating Queries in a Catalog

This topic introduces how to create and edit queries in a catalog as follows:

• Predefining Queries in a Catalog
• Adding Extra Tables to a Query
• Removing Tables from a Query
• Joining Tables in a Query
• Editing the Joins in a Query
• Filtering a Query
  • Filtering with the QBE Format
  • Filtering with the Filter Format
    • Inputting Filter Condition Values Manually
    • Using Sub-queries in the Filter Conditions
• Creating Computed Columns in a Query
• Adding Formula Fields to a Query
• Editing the SQL Statements of a Query
• Previewing a Query

Predefining Queries in a Catalog

1. Open the required catalog, then do either of the following:
   • In the Catalog Manager, expand the data source in which you want to create the query, then select the Queries node or any existing query in the data source and select New Query on the Catalog Manager toolbar, or right-click the Queries node in the data source and select New Query from the shortcut menu.
   • Select Home > New > Query or File > New > Query, specify the data source in which the query will be created, then select OK.
2. In the Enter Query Name dialog, give a name for the query and select OK. The Add Tables/Views/Queries dialog appears.

   

3. Add the resources based on which to create the query.

   If the catalog data source is connected with multiple connections, you can mash up multiple data resources in the query that come from all these connections, including tables, views, synonyms, queries, imported SQLs, imported APEs, stored procedures and user defined data sources. When a query, imported SQL, imported APE, stored procedure or user defined data source is added, it will be added as a single table with all of its columns. When two resources (for example, a table and a view) use the same name, they cannot be added to the query at the same time, and when a table is already contained in a query, you cannot add the table and the query at the same time.

   If the current catalog data source contains JDBC connections, the Show Tables/Views Already Added option is available. Select it if you want to show the tables, views and synonyms in the resource tree in the left box, which have already been added to the right box. You can then add the tables, views and synonyms of the JDBC connections to the query as many times as you want by providing different names for the tables, views and synonyms each time you add them.

4. Select OK. The Query Editor appears, with the selected resources displayed as tables.

   

5. You can select Add Tables on the toolbar to add more data to the query.
6. Select the required columns in each table. To select all, select *. The columns will be displayed in the criteria panel in the lower part of the Query Editor. However for tables from MongoDB connections, you cannot select all columns by selecting *. The PrimaryKey and ForeignKey columns in such table cannot be selected to a query.
7. Join the tables as required.
8. Create filters to narrow down data retrieved to the query.
9. You can add computed columns and formula fields to the query.
10. Select Apply to save the query and then select OK to close the editor.

Besides predefining queries in a catalog from the Catalog Manager, you can also add queries to a catalog from the query resource tree in the Data screen of the report wizard. For example, when you insert a chart in a query based page report, you can select the <New Query...> item in the resource tree to create a query using the Query Editor for the chart.

Adding Extra Tables to a Query

1. In the Query Editor, select Menu > Query > Add Tables/Views/Queries. The Add Tables/Views/Queries dialog appears.

   

2. In the All Tables/Views/Queries box, all the tables, views, synonyms, queries, imported SQLs, stored procedures and user defined data sources in the current catalog data source in which the query is added are listed. Select the resources you want to use in the query and select to add them into the Selected Tables/Views/Queries box. You can mash up multiple data resources from the same catalog data source in a query if you want. When two resources (for example, a table and a view) use the same name, they cannot be added to the query at the same time, and when a table is already contained in a query, you cannot add the table and the query at the same time.

   To remove unwanted resources from the Selected Tables/Views/Queries box, select them and select .

3. If the current catalog data source contains JDBC connections, the Show Tables/Views Already Added option is available. Select it if you want to show the tables, views and synonyms in the resource tree in the left box, which have already been added to the right box. You can then add the tables, views and synonyms of the JDBC connections to the query as many times as you want by providing different names for the tables, views and synonyms each time you add them.
4. Select OK to add the selected resources to the current query. When a query, imported SQL, stored procedure or user defined data source is added, it will be added as a single table with all of its columns unselected to be included.
5. Select Menu > Query > Arrange Table/View/Query to organize the tables.
6. Select the required columns in each table. To select all columns in a table, select * at the top of the table. The columns will be displayed in the criteria panel in the lower part of the Query Editor. However for tables in MongoDB connection, you cannot select all columns by selecting *. The PrimaryKey and ForeignKey columns in each table cannot be selected to a query.

Removing Tables from a Query

1. In the Query Editor, select the table that you want to delete, then do any of the following:
   • Select Delete Table on the toolbar.
   • Select Menu > Query > Delete Table/View/Query.
   • Select the x button at the top right corner of the table.
2. Repeat the above steps to delete other tables. After deleting a table from a query, any joins based on the table will be removed.

Joining Tables in a Query

Based on the Auto Join options that are selected in the Query Editor category of the Options dialog, Logi Report will join the tables in a query automatically. You can also add more joins among these tables manually. Any joins defined between two tables are represented as arrows connecting the key fields from the two tables. Logi Report supports joining tables that come from different connections in the same catalog data source to create distributed joins.

• To join the tables in a query automatically:
  The Auto Join feature enables you to join tables together automatically based upon the following criteria:
  • Foreign Keys
    If there is a column that is defined as foreign key in one table and a primary key in another at the same time, these two tables will be joined together. Not all database systems provide enough information to JDBC drivers to recognize this condition.
  • Primary Keys with Same Names
    If a column is defined as a primary key in one table and appears in another table in the same query, these two tables will be joined together.
  • Same Column Names
    Tables with the same column name will be joined together. This usually results in many joins being not valid. It takes longer to remove the invalid joins than it does to delete the tables. In this case, delete the tables, then add them again to add the joins manually.

  By default all the three auto join criteria are turned on. If you do not want to apply any criteria in a query, in the Query Editor open Menu > Query > Auto Join and clear the criteria you want to turn off. To turn off the criteria for all queries, go to File > Options > Query Editor and clear the criteria you want to turn off in the Options dialog. However if you have specified to use pre-joins in your queries, the Auto Join feature will not take effect.

• To join the tables in a query manually:
  • Position the mouse pointer over the column that will be the source of the join, then select and hold the left mouse button while dragging the join away from the source column to the destination column.
  • Select the columns from two tables while pressing CTRL, then select Menu > Query > Join Columns.

  The join relationship is then established between the tables. When more than one relationship is required between two tables, you can create multiple joins between them.

However when you use the Auto Join feature to join the tables in a query automatically, you often see many joins that are not valid. You can delete unneeded joins by double-clicking the join icon and selecting Delete Join in the Join Options dialog.

Alerting when Cartesian product is used

A Cartesian product is used when tables are included in a query together with no join specifications. You can specify whether to alert when this happens for a query as follows: in the Query Editor select Menu > Query > Current Query Option, then in the Query Options dialog, select or clear Warning When Cartesian Exists and select OK.

For example, Table A has three values: A, B and C. Table B has three values: 1, 2 and 3. Value A matches value 1, value B matches value 2, and so on. This is a specific match. However a Cartesian product could have value A matching with 1, 2 and 3, and value B matching with 1, 2 and 3, and so on. Depending on the data values, Cartesian products can produce a large dataset as unnecessary information will be duplicated. For every record in Table A a record will be created for every record in Table B, thus if Table A has 10 records and Table B has 10 records the result will be a dataset containing 100 rows. However not all Cartesian products are bad so you could use Cartesian product if the result is what you need.

Editing the Joins in a Query

1. In the Query Editor, double-click the join icon in the join line. The Join Options dialog appears.

   

2. To make the join an outer join, select the Outer Join option, then select either the Left, Right or Full radio button if you would like all rows of the left table, right table, or both tables to be retrieved. Regardless of where the tables are placed in the Query Editor, left table is where the arrow starts and right table is where the arrow points.
3. Edit the join condition in the Condition panel according to your requirements.

   Select beside the two text boxes to select a column in the two tables involved in the join, or a parameter or constant level formula in the current catalog data source and select the operator to compose the condition. You can also manually type the column, parameter, or formula name in the text boxes (the input format for parameters and formulas should be @FieldName or :FieldName.). When a parameter is referenced in a join condition, the Ignore Predicate If Parameter Value Is Null setting of the parameter will be ignored. Using parameters in the join conditions can dynamically change query results at runtime. It works similarly as in query filters.

   To add another condition line, select the Add Condition button and define the condition as required. Then from the logic drop-down list, specify the relationship between the two condition lines. You can repeat this to add more condition lines.

   To make some conditions grouped, select them and select the Group button, then the selected conditions will be added in one group and work as one line of conditional expression. Conditions and groups together can be further grouped.

   To take any condition or group in a group out, select it and select Ungroup.

   To adjust the priority of the conditions, select it and select the Up or Down button.

   To delete a condition line, select it and select the Delete button.

4. Select the OK button to accept the changes and close the Join Options dialog.

   Select the Delete Join button if you want to delete the join.

Filtering a Query

You can specify criteria to filter the columns to be retrieved from the database for a query, so that when you build reports on the query, the returned results will be narrowed down.

Filter condition can be composed in both the format of a QBE (Query By Example) and a filter. Using the filter format, you can define the condition on the query and any tables in the query. The overall filter condition applied to a query will include all of them, that is QBE filter AND query filter AND table filter. Always select the SQL button to view your condition and parse the query to ensure the syntax is valid.

Filtering with the QBE format

The criteria panel in the lower part of the Query Editor is for you to filter out some unnecessary records in a query. This filter is in the format of Query by Example (QBE). The advantage of QBE retrieval is that you don't need to learn a query language to frame a query. All the data fields added to a query are shown to you, and all you need to do is to specify the information that restricts the search to the required criteria. Any fields left blank will match everything.

When filtering with the QBE format, you are prompted to type the search criteria into a template resembling the data record. For example, if a column is labeled REGION, and it is a list of all 50 states in the United States. If you only want to see information from California (CA) and New York (NY), in the criteria panel, you can pick out CA and NY by placing their names in the column as below.

You just need to type in NY and CA. Logi Report Designer automatically places ='xx' (equal sign and quotes). See a list of the syntax available:

• Comparison predicates ( =, >, <, >=, <=, <> ).
• BETWEEN predicate (example: BETWEEN 1 AND 100).
• IN predicate (example: NOT IN (1, 3, 5)).
• LIKE predicate (example: LIKE '%apple%').
• NULL predicate (example: IS NOT NULL).

Note that in an Oracle database, if you want to use a Date or DateTime type parameter or a specific date or time to filter the column in a query, you need to apply the to_date() or to_timestamp() function in the filter condition, for example:

Filtering with the Filter Format

Compared with QBE, the filter format provides you with more flexibility with composing the conditions. The expression includes not only the DBFields, but also formulas and parameters. You can also manually type in strings that are supported by the database, for example, you can reference functions predefined in the database like qualifier.functionname (qualifier is optional) in the conditions.

When you use the filter format to filter a query, you can add the filter conditions on both the query and any tables in the query. A filter based on a query is applied as long as the query is used or referenced, while a filter based on a specific table in a query is applied only when the table is queried at runtime.

• To create a filter based on a query, select Menu > Query > Filter in the Query Editor. In the Search Condition dialog specify the filter conditions. The resources available for use are the table columns in the query, the parameters, queries and valid formulas of the query in the same catalog data source as the query, and the special field User Name.
• To create a filter based on a specific table in a query, select the button on the title bar of the table in the Query Editor. In the Table Filter Condition dialog specify the filter conditions. The resources available for use are the columns in the table, the parameters, queries and valid formulas of the table in the same catalog data source as the query, and the special field User Name.

You can compose the filter conditions as follows:

1. Select the Add Condition button to add a condition line.

   

2. In the field text box, specify the field to be filtered. You can either type in the name of the field manually (the input format should be @FieldName or @"Field Name" when the field name contains blank space) or select to specify the field in the Expressions dialog.
3. From the operator drop-down list, set the operator with which to compose the filter expression.
4. In the value text box, select to specify the value of how to filter the field in the Expressions dialog or type the value manually. You can also use sub-queries to narrow down the result, or use the special field User Name or a parameter to filter the query dynamically.
5. Select Add Condition to add more condition lines and define the logic (And or Or) between the condition lines.

   To make some condition lines grouped, select them and select the Group button, then the selected condition lines will be added in one group and work as one line of filter expression. Conditions and groups together can be further grouped. To take any condition or group in a group out, select it and select Ungroup.

   To adjust the priority of the condition lines, select it and select the Up or Down button.

   To delete a condition line, select it and select the Delete button.

6. Select OK to save the condition.

Inputting Filter Condition Values Manually

When you type in the value manually for a condition, you need to pay attention to the following:

• Multiple values should be separated with ",". If "," or "\" is contained in the values, write it as "\," or "\\".
• For String type values
  Quote the values with single quotes.

  Example1: Customers_Country='USA'.

  Example2: Customers_Country in 'Australia','Germany','Mexico'

  The quote marks can be typed by yourself or be added by Logi Report Designer automatically. To have the quote marks automatically added by Logi Report Designer, follow the steps below:

  1. Select Options in the Catalog Manager.
  2. In the Options dialog, switch to the Query Editor category and select Automatically add quotation marks on input values.
  3. Select OK to save the settings.
• For Date type values
  Make sure the format of the value you specify is consistent with that of your database. Filters are done in the RDBMS side and some RDBMS have special requirements for the date format, and you need to make sure the date formats specified in the Data Format tab of the Get JDBC Connection Information dialog are supported by your database.

  In an Oracle database, if you want to use a Date or DateTime type parameter or a specific date or time to filter the fields of a query, you need to use the to_date() or to_timestamp() function in the filter condition, for example:

  

Using Sub-queries in the Filter Conditions

When filtering the fields of a query with the filter format, you are also enabled to use sub-queries to narrow down the result. The following are syntaxes that can be used in sub-queries.

Quantified predicate

A quantified predicate compares a value with a set of values.

• Syntax
  expression----+- = --+- SOME --+-- ( subselect )
                                +- <>  -+  ANY    ---+
                                +- ! = --+  ALL   ----+  
                                +- <   --+
                                +- >   --+
                                +- <=  --+
                                +- ! >  --+
                                +- >=  --+
                                +- ! < --+

  The subselect must specify a single result column and can return any number of values, whether they are null or not.

  • When ALL is specified, the result of the predicate is:
    • True if the result of the subselect is empty, or if the specified relationship is true for every value returned by the subselect.
    • False if the specified relationship is false for at least one value returned by the subselect.
    • Unknown if the specified relationship is not false for any values returned by the subselect and at least one comparison is unknown because of a null value.
  • When SOME or ANY is specified, the result of the predicate is:
    • True if the specified relationship is true for at least one value returned by the subselect.
    • False if the result of the subselect is empty, or if the specified relationship is false for every value returned by the subselect.
    • Unknown if the specified relationship is not true for any of the values returned by the subselect and at least one comparison is unknown because of a null value.
• Examples
  Select qty FROM sales WHERE qty>= ALL (SELECT qty FROM sales)
SELECT BUYERID, ITEM FROM ANTIQUES WHERE PRICE != ANY (SELECT PRICE FROM ANTIQUES);

EXISTS predicate

The EXISTS predicate tests for the existence of certain rows.

• Syntax
  - [ NOT ] EXISTS--(subselect)

  The subselect may specify any number of columns and,

  • The result is true only if the number of rows specified by the subselect is not zero.
  • The result is false only if the number of rows specified by the subselect is zero.
  • The result cannot be unknown.
• Example
  Select DISTINCT pub_name FROM publishers WHERE EXISTS (SELECT * FROM titles WHERE pub_id = publishers.pub_id AND type = 'business')

IN predicate

The IN predicate compares a value with a set of values.

• Syntax
  expression----+-- [NOT] IN --+-- ( subselect )

  In the subselect form, the subselect must identify a single result column and may return any number of values, whether null or not null.

• Example
  Select distinct pub_name FROM publishers WHERE pub_id IN (SELECT pub_id FROM titles WHERE type = 'business')

The following example explains how to apply a subquery when filtering a field:

1. Create a query named mainin in the catalog, add the table Customers and select the following columns: Customers_Customer ID, Customer Name, Customers_City, and Customers_Region.
2. Select Menu > Query > Filter to open the Search Condition dialog.
3. Select the Add Condition button to add a condition line.
4. Select beside the field text box. In the Expressions dialog, select the column Customers_Customer ID, then close the dialog.
5. Select in as the operator from the operator drop-down list.
6. Select beside the value text box. In the Expressions dialog, select the Subquery tab. Select an existing query in the catalog to be the subquery. If you want to edit the selected query, select the Edit Subquery button. To create a new subquery, select the New Subquery button.

   Here, we create a new query named subin, add the table Orders, select the column Orders_Customer ID, and add a condition "Ship Via=Express Delivery" in the Search Condition dialog.

7. Select OK. The subquery subin will then be added into the value text box. Select OK to close the Search Condition dialog.

Now, the subquery subin will be applied to the filter when you build a report that uses the Customers_Customer ID column.

Creating Computed Columns in a Query

Computed columns can be created in a query that is built on tables, views, and synonyms from one JDBC connection only.

1. In the Query Editor, select New Computed Column on the toolbar or select Menu > Column > New Computed Column. The New Computed Column dialog appears.

   

2. Type a name for the column in the Computed Column Name text box.
3. Compose your functions for the column.

   In the lower part of the dialog, there are functions and tables/columns of the query. These are just for your reference. You can specify the expression by yourself in the editing text box, only if the expression can be accepted by your database.

   In addition, the functions in this dialog are not from the Logi Report system. They are from the database you are connected to. If you change your database, some of these functions may no longer exist. For each database, a different set of supported functions will be returned. The following functions will help you with writing an expression.

   • +: Add numbers or fields together in the Expression menu.
   • -: Subtract numbers or fields together in the Expression menu.
   • *: Multiply numbers or fields together in the Expression menu.
   • /: Divide numbers or fields together in the Expression menu.
   • =: Equate fields together.
   • ": Place quotations on long character strings or names that have blanks in them (example: place quotes on values such as "New York" or "Washington DC").
   • ||: Place fields together in the same Expression menu. (Example: "New York" || "Washington DC").
   • (): Place your fields in parentheses.
4. Select OK to create the computed column.

The computed columns added to a query are placed in the criteria panel of the Query Editor together with the table columns in the query. If you want to edit a computed column, you can double-click its name in the criteria panel and then edit in the displayed dialog.

When a computed column is created, it will be added to the SQL of the query. Suppose you have added a computed column named Net Total in a query and the computation is @UNITPRICE * @QUANTITY * (100 - @DISCOUNT) / 100, when you view the SQL statements of the query, you will see the following SQL statement being inserted into the SQL: @UNITPRICE * @QUANTITY * (100 - @DISCOUNT) / 100 AS "Net Total".

Adding Formula Fields to a Query

In addition to the table columns, you can also add formula fields to a query.

1. In the Query Editor, select Add Formula Fields on the toolbar or select Menu > Column > Add Formula Fields. The Add Formula Field dialog appears.

   

2. The formulas in the current catalog data source that are valid for the query are listed in the Formulas box. Choose the required formula in the box and select Add. The formula is then added in the criteria panel of the Query Editor. If the table that contains the columns the selected formula references doesn't exist in the query, this table will be automatically added to the query with the involved columns selected at the same time.
3. Repeat the above step until you have added all the required formulas.
4. Select Close to exit the dialog. The selected formula fields will be added to the SELECT columns in the SQL of the query.

You can replace any formula field added to a query with another one. To do this, double-click its name in the criteria panel and then choose the required field in the Replace Formula dialog.

Editing the SQL Statements of a Query

For a query built on tables, views, and synonyms from one JDBC connection only, its SQL statements can be edited. To do this, in the Query Editor select the SQL button or select Menu > View > Edit SQL, the SQL dialog is then displayed showing the SQL statements used to execute the query like below. You can edit the query statements in the text area as you want. To see whether the statements can be successfully processed by the database, select Execute; to accept the changes you have made, select OK.

Customized SQL mode

When creating a query using the Query Editor, Logi Report allows you to copy and paste your own SQL statements directly in the SQL dialog as explained above. However, because Logi Report query parser recognizes a limited set of the SQL 92 standard, sometimes Logi Report is not able to parse the customized SQL statements. For example, functions such as aggregations using the GROUP BY clause are not supported. In this case, a warning message is displayed after you select the OK button in the SQL dialog asking whether to continue using the customized SQL statements. If you choose Yes, the customized SQL mode is enabled, in which all the functions in the Query Editor will be disabled, and you can only edit the query by modifying the SQL statements in the SQL dialog. You should guarantee the correctness of the customized SQL statements since Logi Report will not parse it although you can check it by using Execute and passing it to the database to check.

While using the customized SQL mode, you can see that the Menu > Query > Customized SQL Mode option of the Query Editor is enabled and selected. By clearing this option, you can go back to use functions of the Query Editor, in which case the customized SQL statements will be lost and the SQL statements generated by Logi Report will be applied instead.

Therefore, if you want to use customized SQL, it is much better to copy your SQL to a text file and import it using the Import SQL feature.

Previewing a Query

1. In the Query Editor, select Preview. The Preview Option dialog appears.

   

2. In the Max Records text box, specify the maximum number of records to be displayed.
3. In the Records per Page text box, specify the number of records displayed on one page.

4. In the Range box, specify the range of the records for previewing.

   • Max Records
     If the option is selected, the Max Records option will take effect.
   • All
     If the option is selected, all the records will be displayed when you preview the query, and the Max Records option will be disabled.
5. Select OK. If the query has parameters, you will be prompted to specify the parameter values.

   The Preview dialog appears showing the records. In the dialog, JTable is used to display the result set. When you open the dialog, the result set is cached, and when you close the dialog, the result set is released.

   

6. Select , , and to browse the records. If the type of the result set is TYPE_FORWARD_ONLY, the last page button will be disabled until you have browsed the last page.
7. To refetch the result set, select Refetch. Select Stop to stop Logi Report from refetching the result set.
8. To print the result set, select Print . If you want to preview the result set before printing, select Print Preview.
9. If you want to set up the page parameters, select Page Setup.
10. Selectthe close button of the dialog to close the Preview dialog.

Previous Topic Next Topic