Akim Demaille wrote:
> Le 10 juil. 2012 à 12:05, Jim Meyering a écrit :
>
>> Akim Demaille wrote:
>>
>>> Le 10 juil. 2012 à 10:14, Jim Meyering a écrit :
>>>
> + case $* in
> +*$nl*) me_=$(printf "$me"|tr "$nl|" '??')
Is it worth testing for both $nl and '|' ?
>>>
>>> I trie
Le 10 juil. 2012 à 12:05, Jim Meyering a écrit :
> Akim Demaille wrote:
>
>> Le 10 juil. 2012 à 10:14, Jim Meyering a écrit :
>>
+ case $* in
+*$nl*) me_=$(printf "$me"|tr "$nl|" '??')
>>>
>>> Is it worth testing for both $nl and '|' ?
>>
>> I tried to keep what you did about $n
Akim Demaille wrote:
> Le 10 juil. 2012 à 10:14, Jim Meyering a écrit :
>
>>> + case $* in
>>> +*$nl*) me_=$(printf "$me"|tr "$nl|" '??')
>>
>> Is it worth testing for both $nl and '|' ?
>
> I tried to keep what you did about $nl in $me (it would break
> the sed command). I used printf, sinc
Le 10 juil. 2012 à 10:14, Jim Meyering a écrit :
>> + case $* in
>> +*$nl*) me_=$(printf "$me"|tr "$nl|" '??')
>
> Is it worth testing for both $nl and '|' ?
I tried to keep what you did about $nl in $me (it would break
the sed command). I used printf, since the \n from echo gave
a spurio
Akim Demaille wrote:
> Here is an updated proposal, that was taking into account
> several comments made last week.
Looks fine, modulo a question and test artifacts:
Thanks!
...
> -warn()
> +# warnf_ FORMAT-STRING ARG1...
> +warnf_ ()
> {
> - for i
> - do
> -echo "$i"
> - done | sed -e "s
Hi all,
Here is an updated proposal, that was taking into account
several comments made last week.
0001-bootstrap-let-warn-be-like-tests-init.sh-s-warn_.patch
Description: Binary data
Akim Demaille wrote:
> Le 6 juil. 2012 à 14:44, Jim Meyering a écrit :
>> No, but if you think it's better, somehow...
>> I meant that warnf would be a very thin wrapper around printf:
>>
>>$ printf '%s\n' a b c
>>a
>>b
>>c
>>
>>> Or should warnf expect a single %s
>>
>> or multiple
On 07/09/2012 05:04 PM, Eric Blake wrote:
>> -warn()
>> +# warnf_ FORMAT-STRING ARG1...
>> +warnf_ ()
>> {
>> - for i
>> - do
>> -echo "$i"
>> - done | sed -e "s/^/$me: /" >&2
>
> This is broken if you invoke './bootstrap', since then $me contains /
> and bootstrap complains:
>
> autorec
On 07/06/2012 07:11 AM, Akim Demaille wrote:
> How about this?
>
> From 96b829fb08217060f20a6a56eddb1a952324fa35 Mon Sep 17 00:00:00 2001
> From: Akim Demaille
> Date: Fri, 6 Jul 2012 15:01:53 +0200
> Subject: [PATCH] bootstrap: let warn be alike tests/init.sh's warn_
>
> Reported by Jim Meyeri
() Jim Meyering
() Fri, 06 Jul 2012 17:49:03 +0200
We want warnf '%s\n' a b c d
to print this (as printf would do):
$me: a
$me: b
$me: c
$me: d
With that added "%s: ", and with "$me" inserted into the
list of arguments, we'd get this:
$me: a
On 07/06/12 06:11, Akim Demaille wrote:
One other tiny nit that is helpful in a die() function:
It isn't necessarily true that "die" always gets invoked
from the main process. If it gets invoked from a subprocess
for any of a variety of reasons, this:
+die() { warn_ "$@"; exit 1; }
won't work.
Thien-Thi Nguyen wrote:
> () Akim Demaille
> () Fri, 6 Jul 2012 16:40:37 +0200
>
>But we have to worry about ',' in $me, which is $0, not
>its basename. Do you want me (Akim, not the variable)
>to change me (the variable, not Akim; not that I am
>immutable) to basename and use /
() Akim Demaille
() Fri, 6 Jul 2012 16:40:37 +0200
But we have to worry about ',' in $me, which is $0, not
its basename. Do you want me (Akim, not the variable)
to change me (the variable, not Akim; not that I am
immutable) to basename and use / in the sed command,
instead of ','?
Akim Demaille wrote:
> Le 6 juil. 2012 à 16:09, Jim Meyering a écrit :
>
>> You're welcome to leave the pipe-to-sed.
>> Worrying about an extra process when processing a diagnostic
>> is probably a pre-optimization anyhow. Besides, then
>> we don't have to worry about whether $me is printf-safe.
>
Le 6 juil. 2012 à 16:09, Jim Meyering a écrit :
> You're welcome to leave the pipe-to-sed.
> Worrying about an extra process when processing a diagnostic
> is probably a pre-optimization anyhow. Besides, then
> we don't have to worry about whether $me is printf-safe.
But we have to worry about
On 07/06/2012 07:09 AM, Jim Meyering wrote:
> You're welcome to leave the pipe-to-sed.
> Worrying about an extra process when processing a diagnostic
> is probably a pre-optimization anyhow.
... unless the diagnostic is something like "program failed"
due to lack of enough system resources to fork
Akim Demaille wrote:
> Le 6 juil. 2012 à 15:21, Stefano Lattarini a écrit :
>
>> Why not shave off the extra forks here?
>>
>>printf "$me: $warnf_format_" "$@" >&2
>>
>> This shouldn't cause problems, unless '$me' contains '%' or '\' characters.
>
> Or end of line. Elsewhere, where I use simil
Le 6 juil. 2012 à 15:54, Stefano Lattarini a écrit :
> On 07/06/2012 03:28 PM, Akim Demaille wrote:
>>
>> Le 6 juil. 2012 à 15:21, Stefano Lattarini a écrit :
>>
>>> Why not shave off the extra forks here?
>>>
>>> printf "$me: $warnf_format_" "$@" >&2
>>>
>>> This shouldn't cause problems,
On 07/06/2012 03:28 PM, Akim Demaille wrote:
>
> Le 6 juil. 2012 à 15:21, Stefano Lattarini a écrit :
>
>> Why not shave off the extra forks here?
>>
>>printf "$me: $warnf_format_" "$@" >&2
>>
>> This shouldn't cause problems, unless '$me' contains '%' or '\' characters.
>
> Or end of line.
Le 6 juil. 2012 à 15:21, Stefano Lattarini a écrit :
> Why not shave off the extra forks here?
>
>printf "$me: $warnf_format_" "$@" >&2
>
> This shouldn't cause problems, unless '$me' contains '%' or '\' characters.
Or end of line. Elsewhere, where I use similar routines,
it is useful to
Akim Demaille wrote:
> Le 6 juil. 2012 à 14:44, Jim Meyering a écrit :
>
>> No, but if you think it's better, somehow...
>> I meant that warnf would be a very thin wrapper around printf:
>>
>>$ printf '%s\n' a b c
>>a
>>b
>>c
>>
>>> Or should warnf expect a single %s
>>
>> or multip
On 07/06/2012 03:11 PM, Akim Demaille wrote:
>
> --- a/build-aux/bootstrap
> +++ b/build-aux/bootstrap
> @@ -1,6 +1,6 @@
> #! /bin/sh
> # Print a version string.
> -scriptversion=2012-07-06.11; # UTC
> +scriptversion=2012-07-06.13; # UTC
>
> # Bootstrap this package from checked-out sources.
>
Le 6 juil. 2012 à 14:44, Jim Meyering a écrit :
> No, but if you think it's better, somehow...
> I meant that warnf would be a very thin wrapper around printf:
>
>$ printf '%s\n' a b c
>a
>b
>c
>
>> Or should warnf expect a single %s
>
> or multiple %-directives:
>
>$ prin
Le 6 juil. 2012 à 14:44, Jim Meyering a écrit :
> No, but if you think it's better, somehow...
> I meant that warnf would be a very thin wrapper around printf:
>
>$ printf '%s\n' a b c
>a
>b
>c
>
>> Or should warnf expect a single %s
>
> or multiple %-directives:
>
>$ prin
On 07/06/2012 02:37 PM, Akim Demaille wrote:
> Le 6 juil. 2012 à 14:24, Jim Meyering a écrit :
>
>>warnf '%s\n' "Error: '$app' version == $inst_ver is too old" \
>> " '$app' version >= $req_ver is required"
>
> Don't you mean warnf '%s\n%s\n'? Or should warnf expect a si
Akim Demaille wrote:
> Le 6 juil. 2012 à 14:24, Jim Meyering a écrit :
>
>> Akim Demaille wrote:
>>> Le 6 juil. 2012 à 13:56, Jim Meyering a écrit :
>>>
Um... this is a good reason to post an adjusted patch.
What I meant was that it is fine to continue to print that diagnostic
on one
Le 6 juil. 2012 à 14:24, Jim Meyering a écrit :
> Akim Demaille wrote:
>> Le 6 juil. 2012 à 13:56, Jim Meyering a écrit :
>>
>>> Um... this is a good reason to post an adjusted patch.
>>> What I meant was that it is fine to continue to print that diagnostic
>>> on one line, but not using echo; u
Akim Demaille wrote:
> Le 6 juil. 2012 à 13:56, Jim Meyering a écrit :
>
>> Um... this is a good reason to post an adjusted patch.
>> What I meant was that it is fine to continue to print that diagnostic
>> on one line, but not using echo; using the new "warn".
>>
>> It should use warn, and warn sh
Le 6 juil. 2012 à 13:56, Jim Meyering a écrit :
> Um... this is a good reason to post an adjusted patch.
> What I meant was that it is fine to continue to print that diagnostic
> on one line, but not using echo; using the new "warn".
>
> It should use warn, and warn should be defined as in init.
Akim Demaille wrote:
> Le 6 juil. 2012 à 12:01, Jim Meyering a écrit :
>
>> Akim Demaille wrote:
>>
>>> It is also very useful to have a single command that issues
>>> several lines. This is the case here for instance:
>>
>> I would debate the "very useful" part ;-)
>> How about "might be nice, in
Le 6 juil. 2012 à 12:01, Jim Meyering a écrit :
> Akim Demaille wrote:
>
>> It is also very useful to have a single command that issues
>> several lines. This is the case here for instance:
>
> I would debate the "very useful" part ;-)
> How about "might be nice, in the rare event someone sees
Akim Demaille wrote:
> Hi Jim!
>
> Le 6 juil. 2012 à 11:43, Jim Meyering a écrit :
>
>> Thanks for the patch.
>>
>> Those all look like improvements, but I'd prefer that
>> you change the name s/stderr/warn/: "stderr" is not normally
>> used as a verb.
>
> OK. I avoided "warn" because I felt it w
Hi Jim!
Le 6 juil. 2012 à 11:43, Jim Meyering a écrit :
> Thanks for the patch.
>
> Those all look like improvements, but I'd prefer that
> you change the name s/stderr/warn/: "stderr" is not normally
> used as a verb.
OK. I avoided "warn" because I felt it would be valid
for it to include a "
Akim Demaille wrote:
> Independently of the other thread we're having about set -e,
> I had refactored bootstrap a bit.
> -8<---
>
> * build-aux/bootstrap (stderr, die): New.
> Use them.
> ---
> build-aux/bootstrap | 94
> ++---
> 1 file changed, 47
Independently of the other thread we're having about set -e,
I had refactored bootstrap a bit.
-8<---
* build-aux/bootstrap (stderr, die): New.
Use them.
---
build-aux/bootstrap | 94 ++---
1 file changed, 47 insertions(+), 47 deletions(-)
diff --g
35 matches
Mail list logo
