Closed wesgraham closed 4 years ago
Merging #183 into develop will increase coverage by
0.27%
. The diff coverage isn/a
.
@@ Coverage Diff @@
## develop #183 +/- ##
===========================================
+ Coverage 51.75% 52.02% +0.27%
===========================================
Files 28 28
Lines 1314 1359 +45
===========================================
+ Hits 680 707 +27
- Misses 529 538 +9
- Partials 105 114 +9
Continue to review full report at Codecov.
Legend - Click here to learn more
Δ = absolute <relative> (impact)
,ø = not affected
,? = missing data
Powered by Codecov. Last update 77fdee3...ba23add. Read the comment docs.
Can we adjust the
import
command, such that when given a file, it imports a geth keyfile instead of a raw private ke
Yes will work on now.
@colin-axner would appreciate your insights on normalizing everything. I kept the ability to import strictly a private key, but now the --file
flag imports a geth-keyfile. Is there a better way to do this?
Can we adjust the
import
command, such that when given a file, it imports a geth keyfile instead of a raw private key file?
I think the default behavior should be using a raw private key file. From geth account import --help
:
Imports an unencrypted private key from <keyfile> and creates a new account.
Prints the address.
Our default behavior should line up with geth imo, since that is probably the primary usage for importing/exporting keys.
Therefore we should add a separate flag such as encrypted
with a short hand like -e
or something which is a bool and then is treated as an encrypted keyfile.
Furthermore, the import command was implemented by someone else who had a need for it, so that suggests to me that importing a raw keyfile was particularly useful
Also from the geth command, so importing/exporting an encrypted keyfile is sugar for copy/paste into/from hidden directory:
Note:
As you can directly copy your encrypted accounts to another ethereum instance,
this import mechanism is not needed when you transfer an account between
nodes.
Therefore we should add a separate flag such as
encrypted
with a short hand like-e
or something which is a bool and then is treated as an encrypted keyfile.
Got it - will add a -e
flag for now. We can also remove this functionality if desired
I was the one who originally implemented this i think. Importing a raw key file was definitely not used. i think we should default to geth's encrypted keyfile
I was the one who originally implemented this i think. Importing a raw key file was definitely not used. i think we should default to geth's encrypted keyfile
I was the one who originally implemented this i think. Importing a raw key file was definitely not used. i think we should default to geth's encrypted keyfile
72 was the original pr. I don't have strong preference on the default, but both file types should be supported.
Ah Got it. sgtm
Few things when testing. Usage gets displayed on error which is annoying. You can set cmd.SilenceUsage = true
in the pre-run ( still want it false while parsing arguments )
Entering the wrong password on export still results "Successfully exported". Created and empty file
When entering the passphrase for the exported key. The second prompt should on say Repeat passphrase, instead of the same exact prompt again, like how importing does.
- Entering the wrong password on export still results "Successfully exported". Created and empty file
I just tested this and wasn't able to recreate the issue. I think the error case added in export.go resolved this problem.
- When entering the passphrase for the exported key. The second prompt should on say Repeat passphrase, instead of the same exact prompt again, like how importing doe
The current flow is to display "Enter passphrase:", then "Enter new passphrase for exported key:", then "Repeat passphrase:" Is this correct?
Fixes #40
Export to file using ./plasmacli keys export