From Tom Murray:

For archive reference/sharing: There are multiple barriers for collecting source code from a dozen locations/product teams (e.g. security, control, network issues, reporting, etc.) throughout a company. But the product teams do not have a problem with sending an archive of their fossology database. Unfortunately the database only contains the analysis and not the actual source code. This is problematic when trying to create a company-wide inventory since the files cannot be interrogated to determine the license language within the context of the file, etc…. Understanding that the source code cannot be copied into the database it would be great for fossology to collect the text of the code/comments that surround (maybe 20 lines before and after the License statements/phrases) for each instance of a suspected license/phrase. This would provide some context for the detected license. And would not require the repository to be co-located with the database. I believe this would provide more portability for sharing the results of the analysis without centralizing the source code collection.

Product teams would provide an archive (pg_dump) to a central inventory database and retain their source code at their location. If further analysis is required then the original product team that owns the repository can run their local copy of fossology which is connected to their source code repository.