[PATCH 3/3] util.c: extend --print-format with %C, %D, %M
Allan McRae
allan at archlinux.org
Thu Jul 21 09:58:09 UTC 2022
On 27/3/22 23:33, Emil Velikov wrote:
> On Sat, 26 Mar 2022 at 15:54, Jelle van der Waa <jelle at vdwaa.nl> wrote:
>>
>> From: Jelle van der Waa <jelle at archlinux.org>
>>
>> Extend print-format with checkdepends, depends and makedepends.
>>
>> Signed-off-by: Jelle van der Waa <jelle at archlinux.org>
>> ---
>> doc/pacman.8.asciidoc | 3 ++-
>> src/pacman/util.c | 49 +++++++++++++++++++++++++++++++++++++++++++
>> 2 files changed, 51 insertions(+), 1 deletion(-)
>>
>> diff --git a/doc/pacman.8.asciidoc b/doc/pacman.8.asciidoc
>> index 8a9294fc..2040f98d 100644
>> --- a/doc/pacman.8.asciidoc
>> +++ b/doc/pacman.8.asciidoc
>> @@ -239,7 +239,8 @@ Transaction Options (apply to '-S', '-R' and '-U')
>> builddate, "%d" for description, "%e" for pkgbase, "%f" for filename,
>> "%g" for base64 encoded PGP signature, "%h" for sha256sum, "%n" for
>> pkgname, "%p" for packager, "%v" for pkgver, "%l" for location, "%r"
>> - for repository, and "%s" for size.
>> + for repository, "%s" for size, "%C" for checkdepends, "%D" for depends
>> + and "%M" for makedepends.
>> Implies '\--print'.
>>
>>
>> diff --git a/src/pacman/util.c b/src/pacman/util.c
>> index 519765f1..cae024e4 100644
>> --- a/src/pacman/util.c
>> +++ b/src/pacman/util.c
>> @@ -405,6 +405,28 @@ char *strreplace(const char *str, const char *needle, const char *replace)
>> return newstr;
>> }
>>
>> +static char *concat_alpm_depends(alpm_list_t *lst)
>> +{
>> + char *depends = NULL;
>> + char *tmp = NULL;
>> + for(alpm_list_t *i = lst; i; i = alpm_list_next(i)) {
>> + alpm_depend_t *dep = i->data;
>> + char *depstring = alpm_dep_compute_string(dep);
>> + if(tmp) {
>> + asprintf(&depends, "%s %s", tmp, depstring);
>> + free(tmp);
>> + } else {
>> + asprintf(&depends, "%s", depstring);
>> + }
>> + tmp = depends;
>> + free(depstring);
>
> Function does 2x allocation for each dependency. Where a bunch of
> packages easily have 20+.
> In case that matters and one wants to micro-optimise - a pre/re
> allocated 4k page should reduce that.
>
> Just throwing the idea - doubt it matters too much to care.
Given how long I took to review this patch, I will not chase this
optimisation. But happy for someone to provide a patch to do this.
Allan
More information about the pacman-dev
mailing list