DataExport/DLL Implementation Options

DataExport/DLL provides two different methods to output data formats: Cell-by-Cell output, which uses only the DLL API to create a file, and File-Based Translation, which uses intermediate disk files to create an output file. This page provides a more detailed descriptionz of the options for implementing DataExport/DLL in your application.

Implementation Overview

The diagram above illustrates the data translation process of DataExport/DLL and two possible implementations. A) The Cell-by-Cell (or field-by-field) output method, which uses the DataExport/DLL Application Programming Interface (API) to initialize an output file, pass data to the file and close it. B) The File-Based Translation method, which uses an Export Definition File (EDF) to define the final output format, a Raw Data File as a data source and a single API call to the DataExport/DLL to perform the translation.

Diagram of DataExport/DLL data translation and output process.

Cell-by-Cell Output Method

The Cell-by-Cell output method uses the DataExport/DLL API to establish a file output format and then communicate each unit of data to the DataExport/DLL. This process is called Cell-by Cell output because it requires calling the DLL once for each cell-or field in a database-that is output.

In a Cell-by-Cell implementation, the calling application begins translation using DXInitTrans. The data format of each column (or field) is then specified with DXDefData. The calling application then begins writing data for a row (or record) to a buffer with the DXPut commands, completing each row with the DXWriteBuffer command.

The process of passing data with the DXPut commands and writing the buffer continues until all rows, or records, have been written. The calling application then completes the output file with the DXCloseFile command.

File-Based Translation Method

The File-Based Translation method uses an Export Definition File (EDF) to define an output format, a Raw Data File to communicate the data and a single API call to initialize translation. This method of output uses the DataExport/DLL as a kind of batch processor.

In a File-Based Translation implementation, the calling application creates an EDF containing information about the type of output file that is desired, the column definitions (or field definitions) and names of the files. The application also writes a Raw Data File in a simple ASCII format to be processed by DataExport/DLL. Finally, the calling application issues a DXLaunchEngine call to begin translation.

File-Based Translation Method Using a DataImport Mask

The File-Based Translation method may also be used with a DataImport for Windows Mask file (MSK) substituted for the EDF and, optionally, a report file or other text file may be substituted for the Raw Data File. Developers wanting to use this method should develop their Mask using DataImport and then follow the instructions for the File-Based method, substituting an MSK for the EDF and, optionally, a text file for the Raw Data File.