1 |
On 09/23/2012 09:39 PM, Mike Frysinger wrote: |
2 |
> On Sunday 23 September 2012 23:53:10 Zac Medico wrote: |
3 |
>> On 09/23/2012 08:44 PM, Mike Frysinger wrote: |
4 |
>>> On Sunday 23 September 2012 22:06:43 Ambroz Bizjak wrote: |
5 |
>>>> + case "$EAPI" in |
6 |
>>> |
7 |
>>> case ${EAPI} in |
8 |
>> |
9 |
>> If somebody manages to corrupt the EAPI with an invalid value containing |
10 |
>> space, then you can get a bash syntax error there if ${EAPI} is |
11 |
>> unquoted. Yeah, it's a user error, but it could confuse them even more |
12 |
>> than they are already if they see the syntax error and think that |
13 |
>> portage is at fault. Also, we might consider that a corrupt EAPI can |
14 |
>> come from a binary package (or /var/db/pkg) enviroment.bz2, and so it |
15 |
>> may not even be the current user who is at fault. |
16 |
> |
17 |
> i don't think that's true. case is sane unlike `test` and `[`. |
18 |
> $ f='a b c'; case ${f} in *) echo yeah;; esac |
19 |
> yeah |
20 |
> -mike |
21 |
|
22 |
Sorry, my mistake. I could have sworn that I was able to trigger a |
23 |
syntax error with this in the past. |
24 |
-- |
25 |
Thanks, |
26 |
Zac |