Use case: when turning a script into a KNIME node, we want to know what inputs will be populated by the framework so that an appropriate input dialog can be created for that node. A PreprocessorPlugin is insufficient because the node dialog is used outside of the module running workflow.
One thought is to add API to the ModuleService to easily create a ModuleRunner, which itself has API for preprocessing. Then the module can be inspected to find unresolved inputs and those inputs used to create the dialog.
This may necessitate more granular requests of Preprocessor plugins - e.g. "build a ModuleRunner using headless preprocessor plugins".
Use case: when turning a script into a KNIME node, we want to know what inputs will be populated by the framework so that an appropriate input dialog can be created for that node. A
PreprocessorPluginis insufficient because the node dialog is used outside of the module running workflow.One thought is to add API to the
ModuleServiceto easily create aModuleRunner, which itself has API for preprocessing. Then the module can be inspected to find unresolved inputs and those inputs used to create the dialog.This may necessitate more granular requests of Preprocessor plugins - e.g. "build a ModuleRunner using headless preprocessor plugins".