Link Products to Time Series to allow creation of product information
This configuration file is used to define the products that will be used by the forecast product information panel. More information on this panel can be found in the User Guide > 32 Forecast Product Information Panel.
Each configured product will contain one or more time series filters to configure which time series should be associated with the product. It is only possible to add product information to forecast time series (external or simulated), so while it is possible to associate a non-forecast time series with a product, there is no use in doing so. The products are organised into categories. View and edit permissions can be given to each product to prevent users from viewing or editing product information on a product they should not have access to. The forecast product information panel and its configuration are available in 2019.02.
An example of a Products.xml file is provided below.
The separate elements are explained below.
All products are organised into categories. There can be several <productCategory> elements. The <productCategory> element will contain <product> elements for each of the products in the category. A ProductCategory element has two attributes:
- id: mandatory element, each category must have a unique id
- name: optional element, display name of the category. When not present, its id will be used instead
When information on a product is displayed, the name of its category is shown as well.
Used to define a product in FEWS. The <product> element has two attributes
- id: mandatory element, each product must have a unique id
- name: optional element, display name of the product. When not present, its id will be used instead
The <product> element has several sub elements which are explained below.
Mandatory element. The expiry time indicates how long product information should be stored and displayed by FEWS. For example, setting it to seven days as in the example above will result in each comment added to the product to be available and stored in the system for seven days. This expiry time is counted from the moment the product information is first created.
Optional element. This element should contain the id of the permission users need to have to be able to view this product's information. A user that does not have this permission will still be able to view the time series, but the forecast product information panel will act as if there is no product associated with the time series if the user does not have view permission for its associated product.
Optional element. this element should contain the id of the permission users need to have to be able to edit and create product information for this product. A user that does not have this permission will still be able to view the product information added for the product by other users (provided they do have view permission). However, they will not be able to create, edit or delete product information on this product.
It is mandatory to configure at least one <timeSeriesFilter> element, there can be several <timeSeriesFilter> element in the same <product> element. The <timeSeriesFilter> elements are used to configure which time series sets the product is associated with. For a time series to be associated with a product, its time series set must match at least one of the time series filters configured for that product.
For example, in the example configuration above the "NWP Short Range" product has a time series filter for time series sets with module instance id "Import_NWP_Shortrange". Any time series set with this module instance id will be associated with the "NWP Short Range" product. These time series sets do not need to match the other time series filter configured for the "NWP Short Range product". Similarly, any time series set with parameter "P.catchment", ensemble "DET" and member "SR" will be associated with the "NWP Short Range" product as well, even if their module instance id does not match the first filter.
Each time series can only be associated with one product. If a time series could be associated with several products, because it matches a configured time series filter for several products, FEWS will assume its associated product is the first matching product it finds. There is no guarantee the order in which the products are checked will remain the same. It is therefore wise to ensure the time series filters are strict enough to only match those time series you wish to associate with that product.