Abstract
With the era of data evolution, enterprises increasingly depend on data utilization tools to import or export data from various data sources. Traditionally, enterprises archive such data into row formats, commonly in CSV files. The flat representation of these files has become an excessive burden to opt the right approach for developing and designing applications that structurally meet business needs. CASE (Computer-Aided Software Environment) tools have been praised by domain experts to build applications by describing their domains in a high abstracted level and automatically generating the appropri-ate implementations. However, these tools lack the appropriate facilities to support efficient and generic bulk data import. In this paper, we present a generic CSV data parser based on EMF (Eclipse Modeling Framework) to automatically map row data into platform-specific models. We define a mapping model which defines the mapping between the CSV files and the target metamodels, and an auxiliary Python script to retrieve the corresponding elements. The experimental evaluation of our parser demonstrates its efficiency to import large CSV files into EMF. In this sense, we aim to increase the adoption of model-based approaches for data-driven use cases by executing bulk and row data import into EMF in an agnostic manner.