Codata tools: portable software for self-describing data files
Dennis Merrill, John McCarthy · eScholarship (California Digital Library) · 1982
This paper describes the Codata tools, a set of programs which read, write, and restructure selfdescribing Codata (common data format) files.These tools manipulate both data and data description, so that the the output of any operation is itself a Codata file.Semantics of results and descriptions of derived Codata files are inherited from descriptions of input Codata files.Following the Software Tools philosophy, the Codata tools are modular -each tool performs a specific limited task.They follow the UNIX and Software Tools conventions of standard input and output.The output of any module can automatically serve as the input of another, and they can be ''pipelined" or "chained" together.Codata tools can be used to extract specified rows and/or columns from a file, to sort a file, to perform relational joins, to perform tabulations by aggregating on common key values, and to perform other operations.The Codata tools are written in RATFOR (a transportable FORTRAN preprocessor), and can be easily adapted to run on any computer where the Software Tools have been implemented.Work is currently under way on substantial enhancements to the Codata file format and the Codata tools to provide for more efficient physical storage formats, more complex data structures, and more extensive, open-ended data description.