After importing a previously deleted key, be able to delete it again.

[jennijuju]

fixes #4621

[arajasek]

It's probably "correcter", but a bit riskier, to instead catch the problem described in #4621 at the "import" stage, by deleting any trash-prefixed files for the newly imported private key.

It's always scary to be deleting private key files, though, so maybe not?

[jennijuju]

It's probably "correcter", but a bit riskier, to instead catch the problem described in #4621 at the "import" stage, by deleting any trash-prefixed files for the newly imported private key.

It's always scary to be deleting private key files, though, so maybe not?

I think there is no harm to keep the trash-key there?

[arajasek]

Approved, but someone else should weigh in on:

It's probably "correcter", but a bit riskier, to instead catch the problem described in #4621 at the "import" stage, by deleting any trash-prefixed files for the newly imported private key.
It's always scary to be deleting private key files, though, so maybe not?

[arajasek]

Might also fix #4306?

[strahe]

When the key is imported again, perhaps the trash-key should be deleted, which does no harm, but it is also unnecessary.when trash-key and wallet-key exist at the same time, how do we understand this situation?

[magik6k]

Ideally we wouldn't overwrite the old key, even if it's basically guaranteed to be the same data

I'd:

• Change Put(..) to put(name, info, retries)
• If retries > 0, append -%d to the name we use for checking
• If it's an existing trash key, call put with retries+1

[arajasek]

LGTM