[pacman-dev] [PATCH] Add -s --sign and -v --verify to repo-add usage message.
There were not added to the end of the list as that would mean
splitting the single gettext string that contains -f and the
"see the man page" message, and I wanted to do as little string
damage as possible.
Signed-off-by: Ray Kohler
On Fri, Mar 25, 2011 at 5:14 PM, Ray Kohler
There were not added to the end of the list as that would mean splitting the single gettext string that contains -f and the "see the man page" message, and I wanted to do as little string damage as possible.
Do string damage, it is a major release. But more importantly, repo-remove --help is going to show these and we might want to try and avoid that- I think the reason they weren't added here.
Signed-off-by: Ray Kohler
--- scripts/repo-add.sh.in | 5 +++++ 1 files changed, 5 insertions(+), 0 deletions(-) diff --git a/scripts/repo-add.sh.in b/scripts/repo-add.sh.in index 9b57ba7..3d6d0c7 100644 --- a/scripts/repo-add.sh.in +++ b/scripts/repo-add.sh.in @@ -74,6 +74,11 @@ repo-remove will update a package database by removing the package name\n\ specified on the command line from the given repo database. Multiple\n\ packages to remove can be specified on the command line.\n\n")" printf "$(gettext "\ +Use the -s/--sign flag to sign the database with GnuPG after updating.\n\n")" + printf "$(gettext "\ +Use the -v/--verify flag to verify the GnuPG signature of the database\n\ +before updating.\n\n")" + printf "$(gettext "\ Use the -q/--quiet flag to minimize output to basic messages, warnings,\n\ and errors.\n\n")" printf "$(gettext "\ -- 1.7.4.1
On Fri, Mar 25, 2011 at 6:16 PM, Dan McGee
On Fri, Mar 25, 2011 at 5:14 PM, Ray Kohler
wrote: There were not added to the end of the list as that would mean splitting the single gettext string that contains -f and the "see the man page" message, and I wanted to do as little string damage as possible.
Do string damage, it is a major release. But more importantly, repo-remove --help is going to show these and we might want to try and avoid that- I think the reason they weren't added here.
The -d and -f flags already suffer from the same problem of showing up is repo-remove --help, when they don't apply there. Maybe I should put conditionals in usage() to tailor it to the right invocation. Really, this usage message is a bit off anyway, as it looks nothing like that of pacman, pacman-key, or makepkg. Thoughts?
On Fri, Mar 25, 2011 at 5:31 PM, Ray Kohler
On Fri, Mar 25, 2011 at 6:16 PM, Dan McGee
wrote: On Fri, Mar 25, 2011 at 5:14 PM, Ray Kohler
wrote: There were not added to the end of the list as that would mean splitting the single gettext string that contains -f and the "see the man page" message, and I wanted to do as little string damage as possible.
Do string damage, it is a major release. But more importantly, repo-remove --help is going to show these and we might want to try and avoid that- I think the reason they weren't added here.
The -d and -f flags already suffer from the same problem of showing up is repo-remove --help, when they don't apply there. Maybe I should put conditionals in usage() to tailor it to the right invocation. Really, this usage message is a bit off anyway, as it looks nothing like that of pacman, pacman-key, or makepkg. Thoughts?
Making it unified and correct for the command it is invoked for is more than welcome- +1 from me. -Dan
On 26/03/11 08:16, Dan McGee wrote:
On Fri, Mar 25, 2011 at 5:14 PM, Ray Kohler
wrote: There were not added to the end of the list as that would mean splitting the single gettext string that contains -f and the "see the man page" message, and I wanted to do as little string damage as possible.
Do string damage, it is a major release. But more importantly, repo-remove --help is going to show these and we might want to try and avoid that- I think the reason they weren't added here.
I think these options are still relevant to repo-remove. Allan
On Fri, Mar 25, 2011 at 5:38 PM, Allan McRae
On 26/03/11 08:16, Dan McGee wrote:
On Fri, Mar 25, 2011 at 5:14 PM, Ray Kohler
wrote: There were not added to the end of the list as that would mean splitting the single gettext string that contains -f and the "see the man page" message, and I wanted to do as little string damage as possible.
Do string damage, it is a major release. But more importantly, repo-remove --help is going to show these and we might want to try and avoid that- I think the reason they weren't added here.
I think these options are still relevant to repo-remove.
-d and -f? If you remove an entry, the whole darn thing goes, regardless of whether there are deltas or files ...files in there. -Dan
On Fri, Mar 25, 2011 at 6:45 PM, Dan McGee
On Fri, Mar 25, 2011 at 5:38 PM, Allan McRae
wrote: On 26/03/11 08:16, Dan McGee wrote:
On Fri, Mar 25, 2011 at 5:14 PM, Ray Kohler
wrote: There were not added to the end of the list as that would mean splitting the single gettext string that contains -f and the "see the man page" message, and I wanted to do as little string damage as possible.
Do string damage, it is a major release. But more importantly, repo-remove --help is going to show these and we might want to try and avoid that- I think the reason they weren't added here.
I think these options are still relevant to repo-remove.
-d and -f? If you remove an entry, the whole darn thing goes, regardless of whether there are deltas or files ...files in there.
That's what checking the code bears out to me as well, but I think Allan's point is that -s and -v actually do apply to repo-remove. While reading through this code, I noticed what looks like a bug - when using repo-remove to remove the very last package ("zip up some emptyness"), the repo won't be signed even if -s was given. Want me to fix that once I'm done with the usage message?
On Fri, Mar 25, 2011 at 5:54 PM, Ray Kohler
On Fri, Mar 25, 2011 at 6:45 PM, Dan McGee
wrote: On Fri, Mar 25, 2011 at 5:38 PM, Allan McRae
wrote: On 26/03/11 08:16, Dan McGee wrote:
On Fri, Mar 25, 2011 at 5:14 PM, Ray Kohler
wrote: There were not added to the end of the list as that would mean splitting the single gettext string that contains -f and the "see the man page" message, and I wanted to do as little string damage as possible.
Do string damage, it is a major release. But more importantly, repo-remove --help is going to show these and we might want to try and avoid that- I think the reason they weren't added here.
I think these options are still relevant to repo-remove.
-d and -f? If you remove an entry, the whole darn thing goes, regardless of whether there are deltas or files ...files in there.
That's what checking the code bears out to me as well, but I think Allan's point is that -s and -v actually do apply to repo-remove.
While reading through this code, I noticed what looks like a bug - when using repo-remove to remove the very last package ("zip up some emptyness"), the repo won't be signed even if -s was given. Want me to fix that once I'm done with the usage message?
Yes, it would make sense to do that. Thanks for the QA you're giving this. -Dan
Move the create_signature() call outside the case of non-empty
databases, so it will be called regardless.
Signed-off-by: Ray Kohler
On Fri, Mar 25, 2011 at 6:54 PM, Ray Kohler
On Fri, Mar 25, 2011 at 6:45 PM, Dan McGee
wrote: On Fri, Mar 25, 2011 at 5:38 PM, Allan McRae
wrote: On 26/03/11 08:16, Dan McGee wrote:
On Fri, Mar 25, 2011 at 5:14 PM, Ray Kohler
wrote: There were not added to the end of the list as that would mean splitting the single gettext string that contains -f and the "see the man page" message, and I wanted to do as little string damage as possible.
Do string damage, it is a major release. But more importantly, repo-remove --help is going to show these and we might want to try and avoid that- I think the reason they weren't added here.
I think these options are still relevant to repo-remove.
-d and -f? If you remove an entry, the whole darn thing goes, regardless of whether there are deltas or files ...files in there.
That's what checking the code bears out to me as well, but I think Allan's point is that -s and -v actually do apply to repo-remove.
How's this for usage output?
ataraxia@asenath:~/code/pacman/scripts $ ./repo-add -h
repo-add (pacman) 3.5.1-73-g72775-dirty
Usage: repo-add [-d] [-f] [-q] [-s] [-v] <path-to-db>
On Fri, Mar 25, 2011 at 6:16 PM, Ray Kohler
On Fri, Mar 25, 2011 at 6:54 PM, Ray Kohler
wrote: On Fri, Mar 25, 2011 at 6:45 PM, Dan McGee
wrote: On Fri, Mar 25, 2011 at 5:38 PM, Allan McRae
wrote: On 26/03/11 08:16, Dan McGee wrote:
On Fri, Mar 25, 2011 at 5:14 PM, Ray Kohler
wrote: There were not added to the end of the list as that would mean splitting the single gettext string that contains -f and the "see the man page" message, and I wanted to do as little string damage as possible.
Do string damage, it is a major release. But more importantly, repo-remove --help is going to show these and we might want to try and avoid that- I think the reason they weren't added here.
I think these options are still relevant to repo-remove.
-d and -f? If you remove an entry, the whole darn thing goes, regardless of whether there are deltas or files ...files in there.
That's what checking the code bears out to me as well, but I think Allan's point is that -s and -v actually do apply to repo-remove.
How's this for usage output?
ataraxia@asenath:~/code/pacman/scripts $ ./repo-add -h repo-add (pacman) 3.5.1-73-g72775-dirty
Usage: repo-add [-d] [-f] [-q] [-s] [-v] <path-to-db>
... repo-add will update a package database by reading a package file. Multiple packages to add can be specified on the command line. Options: -d, --delta generate and add delta for package update -f, --files update database's file list -q, --quiet minimize output -s, --sign sign database with GnuPG after update -v, --verify verify database's signature before update
See repo-add(8) for more details and descriptions of the available options.
Example: repo-add /path/to/repo.db.tar.gz pacman-3.0.0.pkg.tar.gz
ataraxia@asenath:~/code/pacman/scripts $ ./repo-remove -h repo-remove (pacman) 3.5.1-73-g72775-dirty
Usage: repo-remove [-q] [-s] [-v] <path-to-db> <packagename> ...
repo-remove will update a package database by removing the package name specified on the command line from the given repo database. Multiple packages to remove can be specified on the command line.
Options: -q, --quiet minimize output -s, --sign sign database with GnuPG after update -v, --verify verify database's signature before update
See repo-remove(8) for more details and descriptions of the available options.
Example: repo-remove /path/to/repo.db.tar.gz kernel26
At quick glance, looks great to me. The manpage might be the next thing to tackle- find a good way to delineate the two sets of options, as it isn't that clear right now. -v is also missing it appears. -Dan
On Fri, Mar 25, 2011 at 7:26 PM, Dan McGee
On Fri, Mar 25, 2011 at 6:16 PM, Ray Kohler
wrote: On Fri, Mar 25, 2011 at 6:54 PM, Ray Kohler
wrote: On Fri, Mar 25, 2011 at 6:45 PM, Dan McGee
wrote: On Fri, Mar 25, 2011 at 5:38 PM, Allan McRae
wrote: On 26/03/11 08:16, Dan McGee wrote:
On Fri, Mar 25, 2011 at 5:14 PM, Ray Kohler
wrote: > > There were not added to the end of the list as that would mean > splitting the single gettext string that contains -f and the > "see the man page" message, and I wanted to do as little string > damage as possible. Do string damage, it is a major release. But more importantly, repo-remove --help is going to show these and we might want to try and avoid that- I think the reason they weren't added here.
I think these options are still relevant to repo-remove.
-d and -f? If you remove an entry, the whole darn thing goes, regardless of whether there are deltas or files ...files in there.
That's what checking the code bears out to me as well, but I think Allan's point is that -s and -v actually do apply to repo-remove.
How's this for usage output?
ataraxia@asenath:~/code/pacman/scripts $ ./repo-add -h repo-add (pacman) 3.5.1-73-g72775-dirty
Usage: repo-add [-d] [-f] [-q] [-s] [-v] <path-to-db>
... repo-add will update a package database by reading a package file. Multiple packages to add can be specified on the command line. Options: -d, --delta generate and add delta for package update -f, --files update database's file list -q, --quiet minimize output -s, --sign sign database with GnuPG after update -v, --verify verify database's signature before update
See repo-add(8) for more details and descriptions of the available options.
Example: repo-add /path/to/repo.db.tar.gz pacman-3.0.0.pkg.tar.gz
ataraxia@asenath:~/code/pacman/scripts $ ./repo-remove -h repo-remove (pacman) 3.5.1-73-g72775-dirty
Usage: repo-remove [-q] [-s] [-v] <path-to-db> <packagename> ...
repo-remove will update a package database by removing the package name specified on the command line from the given repo database. Multiple packages to remove can be specified on the command line.
Options: -q, --quiet minimize output -s, --sign sign database with GnuPG after update -v, --verify verify database's signature before update
See repo-remove(8) for more details and descriptions of the available options.
Example: repo-remove /path/to/repo.db.tar.gz kernel26
At quick glance, looks great to me.
The manpage might be the next thing to tackle- find a good way to delineate the two sets of options, as it isn't that clear right now. -v is also missing it appears.
Ok, here's the second rev of this patch. Hopefully the code is as nice as the output it produces. I'll give the manpage a look as well.
From 4f8b33a5dcde821f843be2989afa5507621ad1eb Mon Sep 17 00:00:00 2001 From: Ray Kohler
Date: Fri, 25 Mar 2011 18:11:15 -0400 Subject: [PATCH] Clean up repo-add usage message
This now includes -s and -v, tailors itself to the current command,
and is formatted more like that of other pacman commands.
Signed-off-by: Ray Kohler
On Fri, Mar 25, 2011 at 6:36 PM, Ray Kohler
Ok, here's the second rev of this patch. Hopefully the code is as nice as the output it produces. I'll give the manpage a look as well.
Can you resend this as not base64 encoded, and using send-email? As an FYI, you can always comment below the triple dashes '---' in a patch email and that won't get applied in the final patch, instead of above the headers like you did. -Dan
This now includes -s and -v, tailors itself to the current command,
and is formatted more like that of other pacman commands.
Signed-off-by: Ray Kohler
participants (3)
-
Allan McRae
-
Dan McGee
-
Ray Kohler