- Pacman-dev - lists.archlinux.org

[pacman-dev] CVS update of pacman-lib/scripts (makepkg)
by Aaron Griffin 01 Feb '07

01 Feb '07

Date: Wednesday, January 31, 2007 @ 20:59:26 Author: aaron Path: /home/cvs-pacman/pacman-lib/scripts Modified: makepkg (1.36 -> 1.37) Move filelist creation ABOVE the other metafiles to keep them out of the list ---------+ makepkg | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) Index: pacman-lib/scripts/makepkg diff -u pacman-lib/scripts/makepkg:1.36 pacman-lib/scripts/makepkg:1.37 --- pacman-lib/scripts/makepkg:1.36 Wed Jan 31 20:18:22 2007 +++ pacman-lib/scripts/makepkg Wed Jan 31 20:59:26 2007 @@ -861,6 +861,11 @@ fi size=$(du -cb $startdir/pkg | tail -n 1 | awk '{print $1}') +# build a filelist - do this first to keep meta files out of the list +msg "Generating .FILELIST file..." +cd "$startdir/pkg" +tar cvf /dev/null * | sort >.FILELIST + # write the .PKGINFO file msg "Generating .PKGINFO file..." cd $startdir/pkg @@ -914,11 +919,6 @@ have_changelog=1 fi -# build a filelist -msg "Generating .FILELIST file..." -cd "$startdir/pkg" -tar cvf /dev/null * | sort >.FILELIST - # tar it up msg "Compressing package..." cd "$startdir/pkg"

1 0

[pacman-dev] CVS update of pacman-lib/lib/libalpm (package.c)
by Aaron Griffin 01 Feb '07

01 Feb '07

Date: Wednesday, January 31, 2007 @ 20:51:12 Author: aaron Path: /home/cvs-pacman/pacman-lib/lib/libalpm Modified: package.c (1.53 -> 1.54) * !strcmp reads funny (as "not string compare" - switch those to == 0, which has a positive connotation. * Added a nice TODO -----------+ package.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) Index: pacman-lib/lib/libalpm/package.c diff -u pacman-lib/lib/libalpm/package.c:1.53 pacman-lib/lib/libalpm/package.c:1.54 --- pacman-lib/lib/libalpm/package.c:1.53 Tue Jan 30 03:14:11 2007 +++ pacman-lib/lib/libalpm/package.c Wed Jan 31 20:51:12 2007 @@ -268,12 +268,15 @@ RET_ERR(PM_ERR_MEMORY, NULL); } + /* TODO there is no reason to make temp files to read + * from a libarchive archive, it can be done by reading + * directly from the archive */ for(i = 0; archive_read_next_header (archive, &entry) == ARCHIVE_OK; i++) { if(config && filelist && scriptcheck) { /* we have everything we need */ break; } - if(!strcmp(archive_entry_pathname (entry), ".PKGINFO")) { + if(strcmp(archive_entry_pathname (entry), ".PKGINFO") == 0) { /* extract this file into /tmp. it has info for us */ descfile = strdup("/tmp/alpm_XXXXXX"); fd = mkstemp(descfile); @@ -296,10 +299,10 @@ FREE(descfile); close(fd); continue; - } else if(!strcmp(archive_entry_pathname (entry), "._install") || !strcmp(archive_entry_pathname (entry), ".INSTALL")) { + } else if(strcmp(archive_entry_pathname (entry), ".INSTALL") == 0) { info->scriptlet = 1; scriptcheck = 1; - } else if(!strcmp(archive_entry_pathname (entry), ".FILELIST")) { + } else if(strcmp(archive_entry_pathname (entry), ".FILELIST") == 0) { /* Build info->files from the filelist */ FILE *fp; char *fn;

1 0

[pacman-dev] CVS update of pacman-lib/scripts (makepkg)
by dan＠archlinux.org 01 Feb '07

01 Feb '07

Date: Wednesday, January 31, 2007 @ 20:18:22 Author: dan Path: /home/cvs-pacman/pacman-lib/scripts Modified: makepkg (1.35 -> 1.36) removal of unnecessary '\' in integrity check generation ---------+ makepkg | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) Index: pacman-lib/scripts/makepkg diff -u pacman-lib/scripts/makepkg:1.35 pacman-lib/scripts/makepkg:1.36 --- pacman-lib/scripts/makepkg:1.35 Thu Jan 25 20:26:08 2007 +++ pacman-lib/scripts/makepkg Wed Jan 31 20:18:22 2007 @@ -626,7 +626,6 @@ msg "Generating ${integrity_name}s for source files" plain "" ct=0 - newline=0 numsrc=${#source[@]} for netfile in "${source[@]}"; do file=$(strip_url "$netfile") @@ -641,8 +640,7 @@ if [ $ct -eq $numsrc ]; then echo ')' else - echo ' \' - newline=0 + echo fi done plain ""

1 0

[pacman-dev] makepkg.conf - trouble 'upgrading'
by Dan McGee 01 Feb '07

01 Feb '07

When we do roll all this out, everyone is going to get a new makepkg.conf.pacnew file. Our format has changed a bit so we should probably mention this in the install message (e.g. no longer putting export in front of variables, new variables, etc.). One of the new additions is the INTEGRITY_CHECK variable. If this is not present in makepkg,conf, makepkg fails silently as I discovered tonight- when generating it does not generate, and the integrity checks never happen when building. I bring this one up more than anything because isn't this fairly standard and unlikely to be changed by the user? Should this just be at the top of makepkg.conf? It seems like a distribution-wide thing that is standard, and it should join BUILDSCRIPT and PKGEXT at the top of makepkg. -Dan

1 0

[pacman-dev] Removal of bindings
by Jürgen Hötzel 31 Jan '07

31 Jan '07

Hi, I don't know why language bindings generation is part of the pacman-lib CVS repository. Any swig-supported (not only perl/python/java) language binding can be generated from alpm.h and a swig interface file. Real language bindings will also ship higher level function/classes. So i would prefer to remove automated bindings generation. Jürgen

3 3

[pacman-dev] Fun supplementary files
by Aaron Griffin 31 Jan '07

31 Jan '07

I wanted to put this out there. There's a handful of supplementary files that would be nice to have in pacman cvs. One that I can think of off the top of my head, is an 'indent.pro' (see The Fine Manual for details) file for GNU indent (based on the pacman coding guidelines). In addition, it'd be nice to have a guide for translators, so they would know how to send translations (I don't feel qualified to write something like that). I had some more, I'll tack them on with other emails later. This is mainly here for people with free time who want to help out, so feel free to submit anything you feel would be worthwhile. Thanks, aaron

2 1

[pacman-dev] New RC up, getting close!
by Aaron Griffin 31 Jan '07

31 Jan '07

Ok, with Dan formally committing to the repo, we seem to be able to get rid of alot of these bugs fairly quickly. There is very little left in the bug tracker, so as long as we all feel comfortable, we should be able to officially release pacman 3 soon. The newest RC (8) is up at: http://archlinux.org/~aaron/pacman/ Please try it out and report _anything_ and _everything_ you can think of, no matter how small it may seem. Tomorrow I'll setup a semi-formal release schedule, and hopefully we can get this out the door and can move on to pumping out features (3.1). So anyway, it's time we all got excited!

4 6

[pacman-dev] Output questions
by Aaron Griffin 31 Jan '07

31 Jan '07

I have a few pending questions about the current output that I'd like to get some opinions on. Firstly, the spacing in -Ss output - there is an extra newline after each package. This was added a while back upon request, and i personally like it, but it makes the output real huge. Looks like the options are: * Keep the line (-Ss output) * Remove the line (-Qs output) * Add the line between repositories only. Secondly, we have an interesting -Ss "issue" here: $ pacman -Ss ^kernel26$ This should only output (duh) kernel26. However, it also displays kernel26beyond. This is because -Ss ALSO searches the provides=() entries for each package... this is a bit misleading. The options are as follows: * Keep this search and indicate, somewhere, that the package was matched due to a provides=() entry * Don't search provides=() at all So, I'd like some opinions on these two topics... there's probably more somewhere, but I can't think of any.

5 5

[pacman-dev] CVS update of pacman-lib/src/pacman (sync.c)
by Aaron Griffin 31 Jan '07

31 Jan '07

Date: Wednesday, January 31, 2007 @ 03:30:16 Author: aaron Path: /home/cvs-pacman/pacman-lib/src/pacman Modified: sync.c (1.104 -> 1.105) Removed the extra newline based on general opinion --------+ sync.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) Index: pacman-lib/src/pacman/sync.c diff -u pacman-lib/src/pacman/sync.c:1.104 pacman-lib/src/pacman/sync.c:1.105 --- pacman-lib/src/pacman/sync.c:1.104 Wed Jan 31 01:10:22 2007 +++ pacman-lib/src/pacman/sync.c Wed Jan 31 03:30:15 2007 @@ -263,7 +263,7 @@ } indentprint(alpm_pkg_get_desc(pkg), 4); - printf("\n\n"); + printf("\n"); } alpm_list_free(ret); } else {

1 0

[pacman-dev] CVS update of pacman-lib (TODO.aaron lib/libalpm/deps.h)
by Aaron Griffin 31 Jan '07

31 Jan '07

Date: Wednesday, January 31, 2007 @ 03:10:03 Author: aaron Path: /home/cvs-pacman/pacman-lib Modified: TODO.aaron (1.13 -> 1.14) lib/libalpm/deps.h (1.17 -> 1.18) * Forgot the int->enum header checkin * Removed the dep-sorting TODO item - switch the sorting to a topological sort later (I'll let Dan do the CS stuff, heh) --------------------+ TODO.aaron | 6 ------ lib/libalpm/deps.h | 2 +- 2 files changed, 1 insertion(+), 7 deletions(-) Index: pacman-lib/TODO.aaron diff -u pacman-lib/TODO.aaron:1.13 pacman-lib/TODO.aaron:1.14 --- pacman-lib/TODO.aaron:1.13 Wed Jan 31 01:54:28 2007 +++ pacman-lib/TODO.aaron Wed Jan 31 03:10:01 2007 @@ -43,12 +43,6 @@ **** BUGS **** * Removal of conflicts on -A and -U (INCOMPLETE) -* ordering of packages to be installed based on dependencies - i.e. pkga depends=(pkgb pkgc), pkgb depends=(pkgc) - pacman -S pkga pkgb pkgc - > a depends on b and c, move them: pkgb pkgc pkga - > b depends on c, move it: pkgc pkgb pkga - > pacman -S pkgc pkgb pkga * gensync: get all info from packages instead of PKGBUILDs for now, we only need to check the PKGBUILD in case of a 'force' option Index: pacman-lib/lib/libalpm/deps.h diff -u pacman-lib/lib/libalpm/deps.h:1.17 pacman-lib/lib/libalpm/deps.h:1.18 --- pacman-lib/lib/libalpm/deps.h:1.17 Tue Jan 23 22:02:54 2007 +++ pacman-lib/lib/libalpm/deps.h Wed Jan 31 03:10:03 2007 @@ -46,7 +46,7 @@ pmdepmod_t depmod, const char *depname, const char *depversion); int _alpm_depmiss_isin(pmdepmissing_t *needle, alpm_list_t *haystack); -alpm_list_t *_alpm_sortbydeps(alpm_list_t *targets, int mode); +alpm_list_t *_alpm_sortbydeps(alpm_list_t *targets, pmtranstype_t mode); alpm_list_t *_alpm_checkdeps(pmtrans_t *trans, pmdb_t *db, pmtranstype_t op, alpm_list_t *packages); int _alpm_splitdep(char *depstr, pmdepend_t *depend);

1 0