Re: [PATCH] Support for long-options in policycoreutils and checkpolicy (Ticket #1 [1672486])

[Joshua Brindle <method@manicmethod.com>]

Joshua Brindle wrote:
> Guido Trentalancia wrote:
>> Dear Eamon,
>>
>> here are the two "maintenance" patches that I did post earlier this
>> month (along with the new manual pages). They are intended to close
>> Ticket #1 [1672486] that I found open on Tresys pages
>> (http://userspace.selinuxproject.org/trac/ticket/1).
>>
>> I do apologize for not putting the keyword "[PATCH]" in the original
>> message.
>>
>> Here is a summary of what has been changed for policycoreutils:
>>
>> - introduced proper handling of -h, -V options and their respective long
>> formats --help and --version to all binaries that are produced from C
>> code. The same issue is not tackled for Python-based tools (e.g.
>> semanage);
>> - relative manual pages have also been updated accordingly.
>>
>> And here is what has been changed for checkpolicy/checkmodule:
>>
>> - introduced proper handling of -h, -V and the long formats --help and
>> --version to all binaries (checkpolicy/checkmodule);
>> - introduced the handling of long options for some of the other
>> available options;
>> - manual pages have also been updated accordingly (and a few
>> undocumented options have been documented).
>>
>> One of the original two patches (the one against policycoreutils) did
>> not compile cleanly anymore against the current HEAD of SELinux, so I
>> have created an updated one, which is the one attached here.
>>
>> It's just a matter of some really minor issues, but I hope it helps, at
>> least the ticket can now be closed...
>>
>> Kind regards,
>>
>> Guido
>>
>
>
> Thank you for doing this. Unfortunately it looks like your mail client
> decided to base64 encode the attachment.
>
> It is always best to use git to generate patch emails, please see:
> http://andrewprice.me.uk/weblog/entry/generating-patch-emails-with-git
>

I also have a couple comments about this patch. First I'm not so sure 
about the:
"[program name] version [version]."

format. For people to be able to test versions in shell scripts I think 
the output should simply be the version.

I'm not sure what the "libtool" hack thing is all about, why is that 
necessary? In fact I'm not sure what the whole opt_program_name thing is 
about, what are you trying to address here? That change makes the patch 
very messy since you are going through and replacing every occurrence of 
argv[0] with program_name. Typically changes should be separated in 
patches to ease review, I'm not sure what this has to do with adding 
--version to the apps.

Also, why -V rather than -v ?

--
This message was distributed to subscribers of the selinux mailing list.
If you no longer wish to subscribe, send mail to majordomo@tycho.nsa.gov with
the words "unsubscribe selinux" without quotes as the message.