Re: [PATCH 1/4] Allow creation of arbitrary git-shell commands
- From
Greg Brockman <gdb@mit.edu>
- Date
- Jul 21, 2010, 15:11 UTC
- Message-ID
- <AANLkTiktvRmx-2iSVCTvrDYrAAw0kGCtsBhb1mg1ULbx@mail.gmail.com>
- In-Reply-To
- <7vr5iym6jj.fsf@alter.siamese.dyndns.org>
Show 27 quoted lines
>> + user_argv[0] = prog;
>> + execv(user_argv[0], (char *const *) user_argv);
>> + free(prog);
>> + }
>> + free(user_argv);
>> + /*
>> + * split_cmdline modifies its argument in-place, so 'prog' now
>> + * holds the actual command name
>> + */
>> + die("unrecognized command '%s'", prog_cpy);
>> + } else {
>> + /*
>> + * split_cmdline has clobbered prog and printed an
>> + * error message, so print the original
>> + */
>> + die("invalid command format '%s'", prog_cpy);
>
> Hmm, we might want to fix split_cmdline to report the breakage better
> then (perhaps not as part of this patch series). Instead of seeing
>
> error: cmdline ends with \.
> fatal: invalid command format 'foo bar \'.
>
> wouldn't it be nicer to see a single error message:
>
> fatal: invalid command 'foo bar\': cmdline ends with \.
>That would definitely be preferable. If possible, I would rather not do it as part of this patch series, but would be happy to work on fixing split_cmdline afterwards.