Manage Semantic Models
- Sandeep Prakash
- Vera Almeida
1. List/Filter the existent Semantic Models
When entering the Model Editor for the first time, when no semantic models exist yet, the list will be empty.
The user can at any given time create new semantic models, using the available action at the right top to Add New Model.
After the creation of models, a list of the existent semantic models is shown, with search/filter possibility as well as the usual management actions to open/edit and delete each model.
For each model exists an action to Open it in the canvas, providing a graphical representation of that model, and also exists an action to Delete that model (inside the Other actions contextual menu), removing it from the system. Generically clicking in the row of each model, in any place not reserved for a specific action, has the same effect of using the Open action of that model.
On the top of the list, the user has the possibility to search models by their name, which provides help finding a specific model once the list starts to have many models. Only models whose name match the search characters will be presented. When no search characters are introduced, all the existent models are shown.
2. Create a new empty Semantic Model
To create a new empty model, use the Add New Model option available:
The New Model popup will be presented requesting the necessary information to create the model:
The model needs to be given a unique name, and a physical connection to the data needs to be provided. There are two possibilities of how to provide the data connection:
Choose an Existing Data Connection already defined in the server
In this case, the dialog will provide a list of all available connections that the user can use, according to the connections currently created in the Pentaho Server instance and to the user’s permissions to access them.
Fill in the details of a New Data Connection, private to the scope of the model, as it won’t be known or reusable anywhere else
In this case, the dialog presents several fields that need to be filled in in order to stablish the new connection. The user has the opportunity to Test Connection if the information provided is enough to stablish a successful connection to the data.
The certified database types by the plugin for this type of connection are: Oracle, MySQL, Microsoft SQL Server, Postgres and MariaDB (given their respective drivers have been added to the server before start).
When the user creates the model, it is initially created empty and the user is redirected to its edition in the canvas (default) mode of the editor, where the user can start build the model using the drag and drop UI experience.
In alternative the user can use the Advanced mode to build the model, where the xml can be directly edited.
To learn more about how to compose the content of a semantic model using SME plugin, see Create a Semantic Model .
2.1 Adding connection parameters to the model
It is possible to also include additional parameters information together with the model when creating it. This possibility is available both for existing as well as for new connections. To add parameters, other then the chosen data source, just activate the toggle Add parameters? at the bottom of the New Model popup. A section to add parameters will be presented:
To add a new parameter, use the Add+ action available, which will create a line in the Parameters table. When the user puts the focus in the parameter Name input, a list of possible parameters is presented that the user can chose from. This field will still accept custom values from the user, in case the user wants to use other than the suggested ones.
There are some reserved keys the system will not allow the user to use as parameter name, as they are reserved. Reserved names that the user can not use are:
ProviderJdbcJdbcDriversJdbcUserJdbcPasswordCatalogCatalogContentCatalogNameDataSourceDataAccessNotListedCatalog
The user can add any number of parameters wanted, just use the Add+ action repeatedly to add more. Any parameter in the table can be individually removed by using the correspondent Delete parameter action at the end of each row. In case the toggle Add parameters? is deactivated, all previous entered parameters in the table will be cleaned, and removed from the model.
3. Import legacy Analysis Data Sources
If the user has previous existing models in the server, the legacy Analysis Data Sources imported into PUC Managed Data Sources, those models can be imported into the Semantic Model Editor, becoming a semantic model.
Use the Import action available:
The system will present a list of the currently existing Analysis Data Sources published in the Pentaho Server, and the user just needs to choose the one to import.
The name of the model, as well as the connection information with which it has been published into the server, will be kept. Its content will be presented in the canvas for the user to visualize it and change it, if wanted.
All Analysis Data Sources that get imported into the Semantic Model Editor will no longer be presented in the legacy PUC Managed Data Sources UI, and start to be listed in the Semantic Model Editor list of models, behaving from that moment onwards as any other Semantic Model.
All existing reports are expected to continue to work transparently after the Import action, as long as the user does not change the structure and/or name of the model in a way that could disrupt those existing reports.
4. Open/Edit a Semantic Model
In the list of all the available semantic models, to open one particular model the user can use the Open action in the correspondent row, or just click on it in any place. The model will be opened by default in the Canvas mode, where the user can drag available elements to add to the model, or edit the ones already existent inside that particular model displayed in the canvas.
Another alternative to edit the model, once in the canvas is to switch into the XML advanced mode, and directly edit the XML content.
To learn more about how to compose the content of a semantic model using SME plugin, see Create a Semantic Model .
5. Delete a Semantic Model
In the list of all available semantic models, exists an action to Delete in each row:
When the Delete action is used for a given model, the system will present a confirmation message, so that the user confirms or cancels the operation that won’t be possible to revert. If the user confirms the operation, the model is deleted, in case the user cancels the operation no change is performed.
