DILCISBoard / SIARD

SIARD (Software Independent Archiving of Relational Databases) - an open file format for the long-term archiving of relational databases
10 stars 0 forks source link

SIARD should specificy an optional file to have the creation report #53

Open luis100 opened 4 years ago

luis100 commented 4 years ago

When transfering a database information into SIARD, there are sometimes issues and details that should be documented. These issues pertain to possible data losses or transformations that happen in the course of the SIARD creation.

Examples:

In these cases, it would be very important to have a file that documents the process of SIARD creation and list issues or warnings found during the creation. The SIARD specification could specify a location within the SIARD file to keep this file, and also define the format of this documentation.

Suggestions: database.siard/reports/creation_report.pdf

Note: other reports might be allowed, for example when editing the SIARD metadata, we might want to document the change in a new report.

luis100 commented 4 years ago

Note that the objective is that when the archive receives the SIARD file it knows where to look for this information and how to read it.

We could consider specifying a machine readable format (like XML), but that would limit the structure and content of the report. In DBPTK we have a set of information that we keep in this report (like information of the parameters used to access the database, parameters used for the creation of siard, list of collumn data type mappings, list of warning, etc.), but that may be implementation and use-case specific.