Use Cases

This page is about what we have to be able to handle, and what we do not. In the discussions of whether to add or remove a particular feature, we've frequently struggled with the issue of whether the capability added is required, what formats it helps us to describe, and whether these are needed in Version 1.0 of DFDL or not.

Required Formats

• Data as described by Cobol file descriptors, C structs, and other programming language data structures
• CSV - comma or tab separated fields with quoting/escaping conventions for fields containing the separator.
• TBD: specific text formats (need list of various industry formats)

Desirable Formats

(not sure about these. If we can't describe these we need to understand exactly why.)

• ISO CGM - common graphics metafile
• Raw Camera sensor data file formats. E.g., Nikon's famous 14-bit/pixel CCD sensor "raw" format file.
• JPEG
• Screen-scraping - deriving data from fixed format print documents. E.g., PDF forms. (Not really "screens", but this is the term used for this sort of thing.)

Non-required formats

• XML itself - no requirement for DFDL to be able to describe well formed XML itself.
• non-VSAM - these are the formats like VBS (variable blocked spanning). DFDL should be able to describe the various block formats individually, but the virtual data layer which is the concatenation of the block payloads, and a format imposed on that payload, these are all layering concepts we do not need to be able to handle in DFDL 1.0