On 2017-04-04 21:00, Dennis Lee Bieber wrote:
On Tue, 4 Apr 2017 13:23:52 -0400, Dave <dbola...@fastmail.fm> declaimed
the following:


I don't care for the idea of replacing the data file for every save. My
preference would to append to the existing data file - makes more sense.
However, that is not how json works.  So, I'm considering other
alternatives for a data file structure.  Paired data (key and value) is
not really required, but my feeling is that it makes the data file more
robust.  A database seems a little over the top for this situation since
there is no querying of data - just load and save.

        Somebody must eventually have to retrieve stuff from this file for some
use; otherwise it is nothing but an ever-expanding waste of disk space.

        A database, at the very least, means you never have to read the old
contents before adding new data. However, lacking any indication of /what/
the data contains ("dictionary of dictionaries" only indicates a nested
structure at the top level -- but does not indicate if it is possible for
an "append" pass to overwrite a pre-existing subdictionary should the outer
dictionary key be a duplicate).

        Possibly a CSV format would suffice...

Would the 'shelve' module module suffice? (OK, it's not plain text, but neither is a DB.)

--
https://mail.python.org/mailman/listinfo/python-list

Reply via email to