[PATCH] CLI/restore: handle missing keys and values in config data.

[David Bremner]

Daniel Kahn Gillmor <dkg at fifthhorseman.net> writes:

> A few notes about the notmuch config as i'm wrapping my head back around
> it:
>
>  * at least for config data stored in the database, a configuration
>    entry with the value of the empty string is the same as an unset
>    configuration value.  is that right?  is that the semantics we want?
>

That's correct. These are the semantics we inherited from Xapian. At the
time I implimented notmuch_database_config_*, that was the feedback I
got, I guess on the grounds of simplicity. It could be changed; I guess
not many people rely on the current semantics

>
>   * we have no "unset" subsubcommand for "notmuch config", only "set"
>     and "query".  should we have "unset" ?
>

Perhaps. Although there is an existing possibility to clear a value from
the config by "notmuch set foo", which works for both values in the
config file and values in the database.  In a sense this is consistent
with the proposed patch.  Even if you care about the distinction between
unset and empty string, the proposed patch is consistent with this
behaviour; the ugly bit is the use of an empty string internally to
signal unset. Probably this deserves some kind of comment in case we
change the handling of empty strings.

> To be clear, there are situations where it's reasonable to have a config
> variable that you want to be the empty string, and you want that to be
> distinct from "unset".

Yes, I basically agree with that.

FWIW, it doesn't really make sense to me to put any substantial effort
into the database config functionality if the concensus is replace it
with a flat file backend, as you and Carl have been discussing. That's
probably a topic for another thread.