Commits
- Commit:
8ce01796c68dfe0e12b6a1869be69cf3b2cf2584
- From:
- Omar Polo <op@omarpolo.com>
- Date:
add `edit' subcommand to interactively modify an entry
It's more or less equivalent to
$ cd /tmp
$ umask 077
$ plass cat entry > tempfile
$ $EDITOR tempfile
$ plass tee entry < tempfile
$ rm tempfile
but way easier. It's also safer because it creates an unique randomly
named file set up with correct permissions.
Suggested by heph, thanks!
- Commit:
91fcc6e3ddaac1838153b32e3d6d453b445309c0
- From:
- Omar Polo <op@omarpolo.com>
- Date:
make `plass find' case-insensitive
This does not affect `plass cat', `plass tee' etc, only the `find'
sub-command.
Discussed with heph.
- Commit:
c992a1e9b1b3f12475053356a5e9e2f0381211f3
- From:
- Omar Polo <op@omarpolo.com>
- Date:
plass: remove the `gen' sub-command; now there is pwg(1)
- Commit:
4602d277c689ba05915898b55c6f2985045aa6b5
- From:
- Omar Polo <op@omarpolo.com>
- Date:
use /dev/urandom instead of /dev/random
From urandom(4) on OpenBSD
For portability reasons, never use /dev/random. On OpenBSD, it is an
alias for /dev/urandom, but on many other systems misbehaves by blocking
because their random number generators lack a robust boot-time
initialization sequence.
- Commit:
fb70a4127eef7cdd86cde443c9409c8d7f247ab6
- From:
- Omar Polo <op@omarpolo.com>
- Date:
check pattern after stripping store path and extension
now `plass find ^keys' will correctly list only entries that starts
with "keys"
- Commit:
dd204b00777178e93099b89fff9abf7c59cee8de
- From:
- Omar Polo <op@omarpolo.com>
- Date:
slightly better error message for no input in `plass tee'
- Commit:
5eba1883785391436224f0052c8148d61a48e2cb
- From:
- Omar Polo <op@omarpolo.com>
- Date:
plass tee: fix multiple line handling, commit message string
previously, `plass tee' was limited to only one line; now it's
unlimited. While here, fix the commit message too: checking the
existence of the file after creating/overwriting it is stupid.
- Commit:
017e8c818783c5b93ec573b916589df7fe020698
- From:
- Omar Polo <op@omarpolo.com>
- Date:
retire `got' and `tog' subcmd
they're just a glorified `cd'
- Commit:
8bd06def3ffe5c08906a5a03583b6dd6579a8f96
- From:
- Omar Polo <op@omarpolo.com>
- Date:
don't require the entry name in `plass gen -n'
- Commit:
f19c01bc72660b995cc26579cfaeac8a8e1743b2
- From:
- Omar Polo <op@omarpolo.com>
- Date:
sync usage with manpage
- Commit:
d16aedaf996808336a77a18c0695b4fc89d9b14e
- From:
- Omar Polo <op@omarpolo.com>
- Date:
default to gpg, the `gpg2' alias was retired
- Commit:
f9b1eaa1638574d50e6c1c822e4f7d34cc0b30e2
- From:
- Omar Polo <op@omarpolo.com>
- Date:
use the correct operators when comparing numbers
eq/lt/gt are for string-wise comparisons!
- Commit:
4568ada0a6df6f406cadc2d8c201e7f4cc533c1a
- From:
- Omar Polo <op@omarpolo.com>
- Date:
create the path if it doesn't exists in 'mv' too
- Commit:
12ad62fdb46a5c0b920e5edc5c5dcd453e732785
- From:
- Omar Polo <op@omarpolo.com>
- Date:
add --use-agent to the gpg2 mix
- Commit:
945f0333b57b269fdbe5bfab1319b6b61264c268
- From:
- Omar Polo <op@omarpolo.com>
- Date:
adjust writepass so gpg2 don't see the output file
gpg has an annoying "feature": it avoids to overwrite files. If in
interactive mode, "gpg2 -o foo" will ask the user whether to override
foo or not, otherwise will just refuse to work.
not nice.
working around this by making gpg2 write to stdout, and redirecting
stdout to that file. Then restore stdout for sanity :)
While here, also make `writepass' fail if gpg fails.