From Ελεύθερο Λογισμικό / Λογισμικό ανοιχτού κώδικα
Jump to navigation Jump to search

FilesDB — utilities to handle file metadata[edit | edit source]

Data[edit | edit source]

The following data should be stored for each file:

  • path
  • size
  • full path (relative to scan point)
  • a number of hashes (md5, sha-1, xxh32, xxh64, git-sha, ssdeep, sdhash, ...)
  • basename
  • extension
  • dirname
  • Scan identifier

Less useful data that can also be stored:

  • permissions
  • dates (creation, modification, access)
  • owner

Each database should also keep track of all the commands (scans) that were used to add content.

The data model should be extensible (for example, more hashes may be added in later versions), so the data storage should take this parameter into account.

Commands[edit | edit source]

The exact set of commands will be determined later.

However, the following should give a reasonable first approach.

The commands can be implemented as subcommands, like in git.

provide help
display the software version
list all scans in a database
display statistics about the DB contents
scan a location and add to the database
rescan a location and update the database according to added, changed, and removed files
quick-scan a location (ignore hashes) and add to the DB
search for a file, according to criteria

Storage[edit | edit source]

All the information is stored into an SQL database.

Individual SQLite files seem to be the perfect point between functionality and complexity, allowing all operations on data and easy transfer between systems.

All the command-line utilities should accept a "-d|--database <arg>" to specify the location of this file.