Databricks Writer: Feature Type Parameters

To access feature type parameters, click the settings icon on a feature type in the workspace. This opens the Feature Type Parameter Editor.

Tip  To always display the editor in FME Workbench, you can select View > Windows > Parameter Editor.

General

All feature types share similar General parameters, which may include Feature Type Name, Reader or Writer information, and Geometry.

In most Writer Feature Type parameter dialogs, you can also control Dynamic Schema Definitions. Some database formats accept Table or Index Qualifier prefixes on the output table feature type.

These parameters apply only to a selected feature type, not to the entire reader.

Important  If a feature type parameter corresponds to a reader-level parameter, then the reader parameter will be ignored and the feature type parameter will be used.

Table Settings: General

This parameter specifies how features will be written into the destination table. Supported feature operations are described below. Note that the described behavior can be dependent on the selected options, as well as the underlying table properties.

Option

Description

If the Row Does Not Exist

If the Row Exists

Insert

The writer appends a new row to a table using input feature attributes.

The writer creates a new row using input feature attributes.

Not always applicable: if the table does not have a unique key or it has an automatically generated unique key, insertion is always possible.

The database cannot violate its key constraints; therefore, errors can occur on row insertion. For example, if there is a unique key and a user specifies the value with the feature, and the feature already exists, then FME Workbench logs an error. This error might be in the form of a rejected feature, or the database may stop processing altogether.

Update

The writer updates existing row(s) in a table using input feature attributes.

A selection method must be specified in the Row Selection group.

The writer rejects the input feature or logs an error if it is unable to continue.

The writer only changes values of the existing row(s) corresponding to the input feature that differ from the input feature.
Update: Only the Match Columns row selection method is available. If a feature in a transaction matches more than one row in the target table, the transaction will fail.

Delete

The writer deletes an existing row(s) in a table.

A selection method must be specified in the Row Selection group.

The writer rejects the input feature or logs an error if it is unable to continue.

The writer deletes existing row(s) corresponding to the input feature.
Delete: Only the WHERE Clause row selection method is available.

Upsert

The writer appends a new row or updates an existing row(s) using input feature attributes and/or geometries.

A selection method must be specified in the Row Selection group.

The writer creates a new row using input feature attributes and/or geometries.

The writer changes values of the existing row(s) corresponding to the input feature that differ from the input feature.

It leaves other attributes untouched.
Upsert: Only the Match Columns row selection method is available. If a feature in a transaction matches more than one row in the target table, the transaction will fail.

fme_db_
operation

The feature operation will be determined by the attribute fme_db_operation on each input feature.

A selection method must be specified in the Row Selection group.

The value of fme_db_operation will be processed according to the steps below.

Note   The processing steps listed below depend on a format's available Feature Operation options.
  • If the value is null, empty, or missing, it will be treated as an Insert.
  • The value will next be matched to Insert, Update, Upsert, Put, and Delete, case-insensitively.
  • If there is no match, the feature will be rejected.
  • If there is a match, the matched feature operation will be performed on the feature.

Tip: The fme_db_operation attribute will be ignored when Feature Operation is set to Insert, Update, Upsert, or Delete.

 The action depends on the operation; however, in general, if nothing is specified, the value is treated as an Insert. The value is treated as an Insert.
fme_db_operation Only the Match Columns row selection method is available.
When creating a new table, the attribute types map, array, and struct cannot be used.

Controls how the feature type handles destination tables:

  • Use Existing – Write to an existing table If the destination table does not exist, the translation will fail.
  • Create If Needed – Create the destination table if it does not exist.
  • Drop and Create – (This option is not available in all formats.) Drop the destination table if it exists, and then create it. The writer will drop and re-create the table before writing any features to it. Tables will be overwritten when the first input feature is processed. If no features are sent to a feature type, then the corresponding table will not be overwritten.
  • Truncate Existing – (This option is not available in all formats.) If the destination table does not exist, the translation will fail. Otherwise, delete all rows from the existing table.

Row Selection

When inserting into a table, Row Selection is ignored. When updating and deleting from a table (if applicable, based on a format's available Feature Operation options), a condition needs to be specified for selecting which rows to operate on. This parameter group offers two methods to construct the selection condition:

Every table must contain a primary key. To update or delete existing rows in a table, the table must have a primary key composed of one or more columns.

The columns specified in the Match Columns dialog will be used for matching destination rows, and one of the selected rows must be a primary key. All matching rows will be selected for update or delete.

If any feature attributes corresponding to the specified match columns contain null or missing values, the feature will be rejected.

This parameter opens a WHERE Clause Builder. You can also type a WHERE clause inline, without launching the Builder. It is optional to start the clause with the word WHERE.

The WHERE Clause Builder makes it easy for users to reference feature attribute values, destination table columns, and invoke FME functions. The WHERE clause is first evaluated as an FME expression, before being passed onto the destination database.

If the WHERE clause is incorrect or if its evaluation results in failure, the translation will fail. Otherwise, if the WHERE clause passes FME evaluation but it is SQL invalid, the feature will be rejected or the translation will fail.

For advanced users, conditional FME expressions created through the Conditional Value editor can be used to create WHERE clauses.

Tip  You can set the WHERE Clause to an attribute. This supports workspace migration and existing workflows involving fme_where. (Direct support for fme_where has been deprecated.) To advanced users who are accustomed to using fme_where, if Feature Operation is set to Update, Delete, or fme_db_operation, an fme_where attribute that conflicts with Match Columns or WHERE Clause will result in feature rejection.