- Pacman-dev - lists.archlinux.org

[pacman-dev] [GIT] The official pacman repository branch, master, updated. v4.0.0rc2-70-gabfa837
by dan＠archlinux.org 11 Oct '11

11 Oct '11

This is an automated email from the git hooks/post-receive script. It was generated because a ref change was pushed to the repository containing the project "The official pacman repository". The branch, master has been updated via abfa8370c0009e415ef2fa97b96c8b042002d92a (commit) via 8f414cd040aa11957a443ddda42ad14b83733007 (commit) via fb401f569e2d85cf90e45deeeeb28992360d60f7 (commit) via 36f31dd133a6833f641823c3e3c12b0c0edca877 (commit) via cc13f4753211599f1579a714f41f55a8c5d99081 (commit) from 9168977b3b33e1aa3f9d130bbd82be4536684171 (commit) Those revisions listed above that are new to this repository have not appeared on any other notification email; so we list those revisions in full, below. - Log ----------------------------------------------------------------- commit abfa8370c0009e415ef2fa97b96c8b042002d92a Author: Dave Reisner <dreisner(a)archlinux.org> Date: Sun Oct 9 23:03:04 2011 -0400 dload: unhook error buffer after transfer finishes Similar to what we did in edd9ed6a, disconnect the relationship with our stack allocated error buffer from the curl handle. Just as an FTP connection might have some network chatter on teardown causing the progress callback to be triggered, we might also hit an error condition that causes curl to write to our (now out of scope) error buffer. I'm unable to reproduce FS#26327, but I have a suspicion that this should fix it. Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 8f414cd040aa11957a443ddda42ad14b83733007 Author: Lukas Fleischer <archlinux(a)cryptocrack.de> Date: Mon Oct 10 15:22:52 2011 +0200 pactree: Make error message smarter in register_syncs() Our error message used to be very unclear when the configuration file could not be found: $ ./pactree -lsr gtk error: failed to register sync DBs Instead, display an accurate message and include the file name: $ ./pactree -lsr gtk error: config file /usr/local/etc/pacman.conf could not be read Also, move the error message inside register_syncs() to allow for differentiating between different errors that might require a handler in the future. Signed-off-by: Lukas Fleischer <archlinux(a)cryptocrack.de> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit fb401f569e2d85cf90e45deeeeb28992360d60f7 Author: Lukas Fleischer <archlinux(a)cryptocrack.de> Date: Mon Oct 10 14:38:59 2011 +0200 pactree: Add "--config" option This allows for specifying an alternate configuration file path, similar to pacman's "--config" option. Given that there is currently no other way to tell pactree to read from another configuration file (except for patching or symlinking), this seems totally sensible - even if there are plans to refactor and/or replace the standalone configuration file parser. We do not define a short option for the sake of consistency with pacman's set of command line options. Signed-off-by: Lukas Fleischer <archlinux(a)cryptocrack.de> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 36f31dd133a6833f641823c3e3c12b0c0edca877 Author: Allan McRae <allan(a)archlinux.org> Date: Sat Oct 8 19:38:01 2011 +1000 makepkg: strip comments after pkgver/pkgrel when checking value Inline comments after pkgver or pkgrel would cause the sanity checks to fail so remove them before checking the value. Signed-off-by: Allan McRae <allan(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit cc13f4753211599f1579a714f41f55a8c5d99081 Author: Allan McRae <allan(a)archlinux.org> Date: Fri Oct 7 16:40:43 2011 +1000 Format pacsort and vercmp testsuite output Make the output into a single block and add separators at the end so that they do not merge into each other. Signed-off-by: Allan McRae <allan(a)archlinux.org> Signed-off-by: Dan McGee <dan(a)archlinux.org> ----------------------------------------------------------------------- Summary of changes: lib/libalpm/dload.c | 5 ++++- scripts/makepkg.sh.in | 4 ++-- src/util/pactree.c | 20 ++++++++++++++++---- test/util/pacsorttest.sh | 6 ++++-- test/util/vercmptest.sh | 6 ++++-- 5 files changed, 30 insertions(+), 11 deletions(-) hooks/post-receive -- The official pacman repository

1 0

[pacman-dev] [PATCH] pactree: Add "--config" option
by Lukas Fleischer 10 Oct '11

10 Oct '11

This allows for specifying an alternate configuration file path, similar to pacman's "--config" option. Given that there is currently no other way to tell pactree to read from another configuration file (except for patching or symlinking), this seems totally sensible - even if there are plans to refactor and/or replace the standalone configuration file parser. We do not define a short option for the sake of consistency with pacman's set of command line options. Signed-off-by: Lukas Fleischer <archlinux(a)cryptocrack.de> --- src/util/pactree.c | 18 +++++++++++++++--- 1 files changed, 15 insertions(+), 3 deletions(-) diff --git a/src/util/pactree.c b/src/util/pactree.c index 09fe101..5e98f79 100644 --- a/src/util/pactree.c +++ b/src/util/pactree.c @@ -76,6 +76,11 @@ static struct color_choices no_color = { "" }; +/* long operations */ +enum { + OP_CONFIG = 1000 +}; + /* globals */ alpm_handle_t *handle = NULL; alpm_list_t *walked = NULL; @@ -90,6 +95,7 @@ int reverse = 0; int unique = 0; int searchsyncs = 0; const char *dbpath = DBPATH; +const char *configfile = CONFFILE; #ifndef HAVE_STRNDUP /* A quick and dirty implementation derived from glibc */ @@ -154,7 +160,7 @@ static int register_syncs(void) { char line[LINE_MAX]; const alpm_siglevel_t level = ALPM_SIG_DATABASE | ALPM_SIG_DATABASE_OPTIONAL; - fp = fopen(CONFFILE, "r"); + fp = fopen(configfile, "r"); if(!fp) { return 1; } @@ -202,6 +208,8 @@ static int parse_options(int argc, char *argv[]) {"reverse", no_argument, 0, 'r'}, {"sync", no_argument, 0, 'S'}, {"unique", no_argument, 0, 'u'}, + + {"config", required_argument, 0, OP_CONFIG}, {0, 0, 0, 0} }; @@ -211,6 +219,9 @@ static int parse_options(int argc, char *argv[]) } switch(opt) { + case OP_CONFIG: + configfile = optarg; + break; case 'b': dbpath = optarg; break; @@ -263,11 +274,12 @@ static void usage(void) " -c, --color colorize output\n" " -d, --depth <#> limit the depth of recursion\n" " -g, --graph generate output for graphviz's dot\n" + " -h, --help display this help message\n" " -l, --linear enable linear output\n" " -r, --reverse show reverse dependencies\n" " -s, --sync search sync DBs instead of local\n" - " -u, --unique show dependencies with no duplicates (implies -l)\n\n" - " -h, --help display this help message\n"); + " -u, --unique show dependencies with no duplicates (implies -l)\n" + " --config <path> set an alternate configuration file\n"); } static void cleanup(void) -- 1.7.7

2 4

[pacman-dev] [PATCH 1/2] dload: unhook error buffer after transfer finishes
by Dave Reisner 10 Oct '11

10 Oct '11

Similar to what we did in edd9ed6a, disconnect the relationship with our stack allocated error buffer from the curl handle. Just as an FTP connection might have some network chatter on teardown causing the progress callback to be triggered, we might also hit an error condition that causes curl to write to our (now out of scope) error buffer. I'm unable to reproduce FS#26327, but I have a suspicion that this should fix it. Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> --- lib/libalpm/dload.c | 5 ++++- 1 files changed, 4 insertions(+), 1 deletions(-) diff --git a/lib/libalpm/dload.c b/lib/libalpm/dload.c index 33824be..e848b30 100644 --- a/lib/libalpm/dload.c +++ b/lib/libalpm/dload.c @@ -377,8 +377,11 @@ static int curl_download_internal(struct dload_payload *payload, /* perform transfer */ payload->curlerr = curl_easy_perform(curl); - /* immediately unhook the progress callback */ + /* disconnect relationships from the curl handle for things that might go out + * of scope, but could still be touched on connection teardown. This really + * only applies to FTP transfers. See FS#26327 for an example. */ curl_easy_setopt(curl, CURLOPT_NOPROGRESS, 1L); + curl_easy_setopt(curl, CURLOPT_ERRORBUFFER, (char *)0); /* was it a success? */ switch(payload->curlerr) { -- 1.7.7

3 5

[pacman-dev] [PATCH] makepkg: strip comments after pkgver/pkgrel when checking value
by Allan McRae 10 Oct '11

10 Oct '11

Inline comments after pkgver or pkgrel would cause the sanity checks to fail so remove them before checking the value. Signed-off-by: Allan McRae <allan(a)archlinux.org> --- scripts/makepkg.sh.in | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/scripts/makepkg.sh.in b/scripts/makepkg.sh.in index 84221d0..50cf272 100644 --- a/scripts/makepkg.sh.in +++ b/scripts/makepkg.sh.in @@ -1448,7 +1448,7 @@ check_sanity() { ret=1 fi - awk -F'=' '$1 ~ /^[[:space:]]*pkgver$/' "$BUILDFILE" | + awk -F'=' '$1 ~ /^[[:space:]]*pkgver$/' "$BUILDFILE" | sed "s/[[:space:]]*#.*//" | while IFS='=' read -r _ i; do eval i=\"$(sed 's/^$['\''"]$$.*$\1$/\2/' <<< "$i")\" if [[ $i = *[[:space:]:-]* ]]; then @@ -1457,7 +1457,7 @@ check_sanity() { fi done || ret=1 - awk -F'=' '$1 ~ /^[[:space:]]*pkgrel$/' "$BUILDFILE" | + awk -F'=' '$1 ~ /^[[:space:]]*pkgrel$/' "$BUILDFILE" | sed "s/[[:space:]]*#.*//" | while IFS='=' read -r _ i; do eval i=\"$(sed 's/^$['\''"]$$.*$\1$/\2/' <<< "$i")\" if [[ $i = *[[:space:]-]* ]]; then -- 1.7.7

1 0

[pacman-dev] [PATCH] Format pacsort and vercmp testsuite output
by Allan McRae 10 Oct '11

10 Oct '11

Make the output into a single block and add separators at the end so that they do not merge into each other. Signed-off-by: Allan McRae <allan(a)archlinux.org> --- test/util/pacsorttest.sh | 6 ++++-- test/util/vercmptest.sh | 6 ++++-- 2 files changed, 8 insertions(+), 4 deletions(-) diff --git a/test/util/pacsorttest.sh b/test/util/pacsorttest.sh index ff82a53..9d52d69 100755 --- a/test/util/pacsorttest.sh +++ b/test/util/pacsorttest.sh @@ -40,10 +40,11 @@ runtest() { if ! type -p "$bin"; then echo "pacsort binary ($bin) could not be located" + echo exit 1 fi -echo "Beginning pacsort tests" +echo "Running pacsort tests..." # BEGIN TESTS @@ -93,11 +94,12 @@ runtest "$separator" "$separator_reverse" "really long input, sort key, separato #END TESTS -echo if [[ $failure -eq 0 ]]; then echo "All $total tests successful" + echo exit 0 fi echo "$failure of $total tests failed" +echo exit 1 diff --git a/test/util/vercmptest.sh b/test/util/vercmptest.sh index 6b4bcbc..04b841f 100755 --- a/test/util/vercmptest.sh +++ b/test/util/vercmptest.sh @@ -64,10 +64,11 @@ runtest() { if ! type -p "$bin"; then echo "vercmp binary ($bin) could not be located" + echo exit 1 fi -echo "Beginning vercmp tests" +echo "Running vercmp tests..." # BEGIN TESTS @@ -145,11 +146,12 @@ runtest 1:1.1 1.1 1 #END TESTS -echo if [[ $failure -eq 0 ]]; then echo "All $total tests successful" + echo exit 0 fi echo "$failure of $total tests failed" +echo exit 1 -- 1.7.7

1 0

[pacman-dev] Improvements to the Norwegian Bokmål (nb) translation
by Alexander Rødseth 10 Oct '11

10 Oct '11

Hi, I felt the Norwegian Bokmål translation could need a little touch up. Attaching nb.patch that can be applied with "git apply". -- Cordially, Alexander Rødseth Arch Linux Trusted User (xyproto on IRC, trontonic on AUR)

2 3

[pacman-dev] [PATCH] PKGBUILD.vim: fix epoch syntax highlighting
by lolilolicon 09 Oct '11

09 Oct '11

--- contrib/PKGBUILD.vim | 3 +++ 1 files changed, 3 insertions(+), 0 deletions(-) diff --git a/contrib/PKGBUILD.vim b/contrib/PKGBUILD.vim index 1814c22..4d69e4d 100644 --- a/contrib/PKGBUILD.vim +++ b/contrib/PKGBUILD.vim @@ -220,6 +220,9 @@ hi def link pb_k_pkgrel pbKeywords hi def link pbIllegalPkgdesc Error hi def link pb_k_desc pbKeywords +hi def link pbIllegalEpoch Error +hi def link pb_k_epoch pbKeywords + hi def link pbIllegalUrl Error hi def link pb_k_url pbKeywords -- 1.7.6.4

1 0

[pacman-dev] Tested pacman 4.0.0rc2-1 - the results
by Daniel Nilsson 08 Oct '11

08 Oct '11

Hi. I've tested pacman 4.0.0rc2-1 on a fresh arch netinstall in virtualbox to test the new signature checking. I followed the instructions on this page (the rc2 announcement): http://mailman.archlinux.org/pipermail/arch-dev-public/2011-September/02163… First I did a full system upgrade with: # pacman -Syyu then I installed pacman rc2 with: # pacman -U pacman-4.0.0rc2-1-i686.pkg.tar.gz install finished without errors. Then I did: # pacman-key --init which yielded the following: "gpg: /etc/pacman.d/gnupg/trustdb.gpg: trustdb createdgpg: no ultimately trusted keys foundgpg: Generating pacman keychain master key... Not enough random bytes available. Please do some other work to givethe OS a chance to collect more entropy! (Need 284 more bytes)" I waited for a while but the command never finished. I tried to reboot, ran "pacman-key --init" once again and got the same message as above but with "(Need 282 more bytes)" instead.Waited for a while to see if it would finish but there was no progress. I hope this feedback will help pacman development. Please let me know if I should provide more info. (I've not tested the x86-64 version yet).

2 1

[pacman-dev] [GIT] The official pacman repository branch, master, updated. v4.0.0rc2-65-g9168977
by dan＠archlinux.org 06 Oct '11

06 Oct '11

This is an automated email from the git hooks/post-receive script. It was generated because a ref change was pushed to the repository containing the project "The official pacman repository". The branch, master has been updated via 9168977b3b33e1aa3f9d130bbd82be4536684171 (commit) via 22abe27bfa4f4e6c3ce3dfc345bb45a1f0666ab8 (commit) via 4928f9edc371a4e4f90f329c9908f00cdbfe5bfb (commit) via 273cc66a8a3173e41b68fd07a0944887a1eb80b9 (commit) via 7e5bbf0387203a078bf901d730f063f10c7c1f2c (commit) via 3f937c8404b54efa34c4774da6d33c352c51aad9 (commit) via 524847145d3977b8eed6203ae280906aaceca28b (commit) from a8ca9b93f8efbbca2311cb0b99d9bd31c1636527 (commit) Those revisions listed above that are new to this repository have not appeared on any other notification email; so we list those revisions in full, below. - Log ----------------------------------------------------------------- commit 9168977b3b33e1aa3f9d130bbd82be4536684171 Author: Dan McGee <dan(a)archlinux.org> Date: Thu Oct 6 00:39:18 2011 -0500 Update NEWS for 4.0.0 release Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 22abe27bfa4f4e6c3ce3dfc345bb45a1f0666ab8 Author: lolilolicon <lolilolicon(a)gmail.com> Date: Thu Oct 6 14:14:46 2011 +0800 bacman: use globbing to get local package db path The original code- pkg_dir="$(echo $pac_db/$pkg_name-[0-9]*)" is problematic in several ways: - $pac_db and $pkg_name should be quoted, obviously. - It assumes pkgver always starts with an integer, while in fact it just can't contain ':' and '-'. Counterexample: the code breaks on lshw B.02.15-1. - It assumes there are no more than one directory matching the pattern. While this should be the case if everything works perfectly, it certainly relies on external conditions. Counterexample: if the local db contains two packages named foo and foo-3g, even if everything else is perfect, the code will match two directories. Don't make assumptions, use what is known. Signed-off-by: lolilolicon <lolilolicon(a)gmail.com> commit 4928f9edc371a4e4f90f329c9908f00cdbfe5bfb Author: lolilolicon <lolilolicon(a)gmail.com> Date: Thu Oct 6 14:14:45 2011 +0800 bacman: add tar.Z package creation support bacman should support whatever makepkg does as PKGEXT. Also remove obsolete $EXT variable. Signed-off-by: lolilolicon <lolilolicon(a)gmail.com> commit 273cc66a8a3173e41b68fd07a0944887a1eb80b9 Author: lolilolicon <lolilolicon(a)gmail.com> Date: Thu Oct 6 14:14:44 2011 +0800 bacman: small code cleanup This includes: - Quoting fixes. - Drop deprecated mktemp option -p. - Set extglob nullglob shell options at the top. - Use extended globbing instead of regex to match %HEADER% in pacman db. Signed-off-by: lolilolicon <lolilolicon(a)gmail.com> commit 7e5bbf0387203a078bf901d730f063f10c7c1f2c Author: lolilolicon <lolilolicon(a)gmail.com> Date: Thu Oct 6 14:14:43 2011 +0800 bacman: bashify using [[ ]] and (( )) Another style change. The [[ expression ]] form is particularly cleaner, safer and more powerful than the [ expression ] form. Signed-off-by: lolilolicon <lolilolicon(a)gmail.com> commit 3f937c8404b54efa34c4774da6d33c352c51aad9 Author: lolilolicon <lolilolicon(a)gmail.com> Date: Thu Oct 6 14:14:42 2011 +0800 bacman: indent code using TAB As every piece of code in the whole project uses TAB as indentation character, bacman shouldn't be an exception. Signed-off-by: lolilolicon <lolilolicon(a)gmail.com> commit 524847145d3977b8eed6203ae280906aaceca28b Author: Dan McGee <dan(a)archlinux.org> Date: Wed Oct 5 23:43:21 2011 -0500 Enable CheckSpace by default in default config file This will have to be picked up downstream of course, but addresses FS#25684 now that this is a lot faster in 4.0 than it was in the original 3.5 implementation. Also make curl the first XferCommand listed, as we are moving away from any other download program at this point. Signed-off-by: Dan McGee <dan(a)archlinux.org> ----------------------------------------------------------------------- Summary of changes: NEWS | 71 +++++++++++ contrib/bacman.in | 333 ++++++++++++++++++++++++++-------------------------- etc/pacman.conf.in | 6 +- 3 files changed, 241 insertions(+), 169 deletions(-) hooks/post-receive -- The official pacman repository

1 0

[pacman-dev] [PATCH 1/6] bacman: get local package db path the correct way
by lolilolicon 06 Oct '11

06 Oct '11

Use globbing and store the expansion results in an array. The [epoch:]pkgver part can start without an integer, e.g. lshw is currently at version B.02.15. extglob is just good enough to match pkgname-[epoch:]pkgver-pkgrel. Set nullglob globally. Signed-off-by: lolilolicon <lolilolicon(a)gmail.com> --- contrib/bacman.in | 16 ++++++++++------ 1 files changed, 10 insertions(+), 6 deletions(-) diff --git a/contrib/bacman.in b/contrib/bacman.in index fe13e5b..b51a4a3 100755 --- a/contrib/bacman.in +++ b/contrib/bacman.in @@ -23,6 +23,9 @@ readonly progname="bacman" readonly progver="0.2.1" +shopt -s extglob +shopt -s nullglob + # # User Friendliness # @@ -90,8 +93,7 @@ pkg_dest="${PKGDEST:-$PWD}" pkg_pkger=${PACKAGER:-'Unknown Packager'} pkg_name="$1" -pkg_dir="$(echo $pac_db/$pkg_name-[0-9]*)" -pkg_namver="${pkg_dir##*/}" +pkg_dir=("$pac_db/$pkg_name"-+([^-])-+([0-9])) # # Checks everything is in place @@ -101,7 +103,11 @@ if [ ! -d "$pac_db" ] ; then exit 1 fi -if [ ! -d "$pkg_dir" ] ; then +if [ ${#pkg_dir[@]} -gt 1 ]; then + echo "ERROR: ${#pkg_dir[@]} entries for package ${pkg_name} found in pacman database" + printf "%s\n" "${pkg_dir[@]}" + exit 1 +elif [ ${#pkg_dir[@]} -eq 0 -o ! -d "$pkg_dir" ]; then echo "ERROR: package ${pkg_name} not found in pacman database" exit 1 fi @@ -109,6 +115,7 @@ fi # # Begin # +pkg_namver="${pkg_dir##*/}" echo Package: ${pkg_namver} work_dir=$(mktemp -d -p /tmp) cd "$work_dir" || exit 1 @@ -275,9 +282,6 @@ esac pkg_file="$pkg_dest/$pkg_namver-$pkg_arch${PKGEXT}" ret=0 -# when fileglobbing, we want * in an empty directory to expand to -# the null string rather than itself -shopt -s nullglob # TODO: Maybe this can be set globally for robustness shopt -s -o pipefail bsdtar -cf - $comp_files * | -- 1.7.6.4

3 31