Directory_Structure

We will follow a directory structure along the lines of the UNIX file system:

The contents of each directory is described below:

• cdbs The root directory of all the CDBS software. We will eventually try to have this directory as /usr/local/cdbs. For now, on the test system, it will be /ralph/data2/cdbs_d/code_d/cdbs.
• util The location of utilities that could be used for example to create makefiles.
• sql The location of all files written in sql. These include the files for creating tables, triggers, stored procedures, and indices.
• include This directory holds the include files that will be shared among CDBS sources. Include files used only for one source should be kept with the source.
• lib The location of all binary files for libraries. The source code for the libraries is kept under the src directory.
• src The root directory to all source code. There should be a separate subdirectory for each CDBS task or library. Each task subdirectory will contain all include files that are specific to the task, as well source, and make files.
• bin Contains the executables for CDBS. They should be links to the executable in the source directory.
• doc Contains the documentation for CDBS in latex.
• data This directory (cdbs/data) holds the data files for use by CDBS tools, such as the catalog file and template files.
• test The root directory for running automatic regression tests.
• data This directory (cdbs/test/data) contains the data for running regression tests, such as reference files.
• script Contains the script files for running regression tests. There should be one script per CDBS task.

Deliveries should consist of a tar file that will extract into the directories indicated above. The extraction will run from the cdbs directory. Libraries should be delivered by the supplier (e.g., B. Simon), rather than users of libraries.