Table Definition Dialog

Table definitions and the table definition dialog underpin all the tabular interfaces - exchanging data with IBM DOORS, importing and exporting CSV files, and publishing tables directly to Microsoft Word. A common approach and a shared set of table definitions is used by all of these mechanisms enabling a "define once, reuse many times" mentality.

The table definitions can be as simple or as complex as desired. For example, one could imagine a basic requirements table consisting of four columns - name, number, description, and children (targets of the refined by relationship). It could also be as complex as a table of components with separate rows for each function allocated to a given component, the description of that function, the inputs / outputs / triggers corresponding to those functions in separate rows, attributes relating to those items, and so on. All of the tables are defined and selected using the table definition dialog as part of the import and export wizards.

Table definitions become part of the project definition (like icon templates, sort blocks, and many other utility objects). They are automatically exported as part of the project file, and using a custom export, they can be exported separately for sharing with other projects and other users.

Table Column Types

Tables are comprised by a series of columns (also loosely referred to as cells). CORE supports the following column types:

Column Type	Output Behavior	Input Behavior
Attribute	Outputs the attribute value in its string form.	Parses the attribute value from its string form.
Class	Outputs the class name (not the alias).	Parses the class from the corresponding name. This cell can then be used as a source cell for resolving an element name into an element.
Class Alias	Outputs the class alias.	Parses the class from the corresponding alias (not the name). This cell can then be used as a source cell for resolving an element name into an element.
Element	Outputs the element name.	Attempts to resolve the element name into an element object. Can use a Class or Class Alias column as a source or can use the specified folder during an import process.
Element ID	Outputs the element ID (a GUID).	Attempts to resolve the element ID into an element object. Since the ID is unique, no source column is required.
Folders	Outputs the path names (one per line) of the folders that contain the source element.	Parses the folders from the path names (one per line). Adds the source element to all of the specified folders.
Parameter	Outputs the parameter value followed by the parameter units (if specified).	Parses the parameter value from its string form. If the parameter type uses units, everything after the first space in the string is treated as units. Sets this as the parameter for the source element.
Parameter Units	Outputs the parameter units for the specified parameter.	Parses the parameter units. Sets this as the units for the specified parameter for the source element.
Parameter Value	Outputs the parameter value in its string form.	Parses the parameter value from its string form. Sets this as the parameter value for the source element.
Relationships	Outputs the relationship objects (one per line). The sort block determines the representation shown in this column. [Unless you need to outputs a relationship attribute in a dependent column, use the Targets column type instead.]	Parses each line to resolve the representation into an element and assigns that element as a target of the specified relationship for the specified element.
Targets	Outputs the target elements (one per line). The sort block determines the representation shown in this column.	Parses each line to resolve the representation into an element and assigns that element as a target of the specified relationship for the specified element.
Unused	Outputs an empty column.	Ignores the column value. The Unused column type is useful if the source data has a column that is not used in CORE.

The Table Definition Dialog

The table definition dialog provides a quick mechanism to access existing table definitions or create new table definitions on the fly. It's applicable for reading input and generating output.

The main body of the dialog consists of a series of banded entries, each entry representing a column specification. Additional entries can be added to the bottom of the list by using the plus button at the top of the dialog. Each entry specifies:

Position. The order (left to right) that the columns should appear in. As you select a new value from the drop-down, the entries immediate re-sort into the new order.
Type. The column type (described above). When the column type is changed, the definition, sort block, header, and output as single cell fields update accordingly.
Definition. The specific attribute, parameter, or relation definition to use (depending upon the column type). In the case of a Relationships or Targets column, the first entry in the list is (multiple values). Selecting this enables the "more" button beside the drop down enabling you to specify multiple relationships / targets in the same cell (appropriate if you are trying to combine all items used by a function in a single column).
Sort Block. For Element, Relationships, and Targets columns, the sort block that will be used to both sort and represent the data. The sort block drop-down provides access to all sort blocks defined for the project.
Based on. The source object for the cell. Every cell type requires an object to act upon (generally an element specification). By specifying the source object, the tables can become arbitrarily complex. In the case shown above, columns 1 through 5 operate on "Data" (the element passed in from the data selection in the wizard). Column 6 operates on column 5 allowing you to output the description of the targets of the basis of relationship.
Header. Text value to output in the first row. In an export operation, this is pre-populated based upon the column type and description, but you can change it as desired.
Output as Single Cell. For column types that include multiple values (targets of a relationship, folders associated with an element), should the entire collection be output in a single cell. If the column will not be used as source data for other columns in the table, it may or may not make sense to output the values as a single cell. If multiple values are output as a single cell, each value is placed on a separate line within the cell.
Repeat Values. When generating tables, it is quite common to generate multiple rows based upon a single value. For example, if column 1 is a requirement, column 2 is the requirement's children, and column 3 is the description for the child requirement, you will likely have many child rows. In this situation, if this option is enabled for column 1, the requirement name will be repeated for every row. If this option is unchecked, the requirement name will be output the first time only. Subsequent rows will contain a blank cell in this column.
Show Column. In most cases, we want columns to show up in the resultant table. Sometimes, we need a column to serve as source data for another column, but we don't actually want to display the source column. Unchecking this option will hide the column in the resultant table.
Delete Button. Clicking the X at the end of any entry will remove that entry from the list.

The header at the top of the dialog contains three critical items:

Stored Definitions. The definitions drop-down allows you to select an existing table definition to use - either exactly as it stands or as a basis for customization. If you select a table definition from this drop-down, you will be warned before it overwrites your table definition.
- Once you have the table definition the way you like it, you can use the save button to store it for reuse in the future. You will be prompted for a name. Table definition names must be unique, so if you specify an existing name it will overwrite the existing definition.
- If you wish to discard a stored definition, you can use the delete button next to the save button.
Sort Block. In the case of an export wizard, the sort block will be applied to the source data to determine the order in which the data is processed. For example, if operating on a package, the sort block will order the packaged elements correctly.
Filter. In the case of an export wizard, the filter will be applied to the source data before it is output. Any elements that do not pass the filter will not be output. For example, if you wish to output a table of open concerns, you could filter the Concern folder accordingly. This list is populated with the stored filter definitions, so you must save any custom filters before using them as part of an export operation.

Though western culture often thinks left-to-right, the columns of a table can be placed in any order desired. In other words, it is valid for column 3 to use column 7 as its source data. CORE processes each row in an order calculated from the source data cells, not the left-to-right order.