November 2013 - Pacman-dev - lists.archlinux.org

[pacman-dev] [PATCH] Fix minor grammatical issues in doc/index.txt
by Jason St. John 08 Nov '13

08 Nov '13

Signed-off-by: Jason St. John <jstjohn(a)purdue.edu> --- This is a resubmit based on Allan's feedback on the ML here: https://mailman.archlinux.org/pipermail/pacman-dev/2013-October/018140.html doc/index.txt | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/doc/index.txt b/doc/index.txt index 9916eab..3c4339e 100644 --- a/doc/index.txt +++ b/doc/index.txt @@ -21,22 +21,22 @@ option) with a master server through the use of package databases. Prior to this, packages would have to be installed manually using the '\--add' and '\--upgrade' operations. -Version 3.0 was the switch to a two-part pacman- a backend named libalpm -(library for Arch Linux Package Management), and the familiar pacman frontend. +Version 3.0 was the switch to a two-part pacman -- a back-end named libalpm +(library for Arch Linux Package Management) and the familiar pacman front-end. Speed in many cases was improved, along with dependency and conflict resolution being able to handle a much wider variety of cases. The switch to a library-based program should also make it easier in the future to develop alternative front ends. Version 4.0 added package signing and verification capabilities to the entire -makepkg/repo-add/pacman toolchain via GNUPG and GPGME. +makepkg/repo-add/pacman toolchain via GnuPG and GPGME. Documentation ------------- -Manpages -~~~~~~~~ -There are several manpages available for the programs, utilities, and +Man Pages +~~~~~~~~~ +There are several man pages available for the programs, utilities, and configuration files dealing with pacman. * linkman:PKGBUILD[5] @@ -194,7 +194,7 @@ link:https://mailman.archlinux.org/pipermail/pacman-dev/[view the archives]. Source Code ~~~~~~~~~~~ -Development of pacman is currently done in GIT. The central repository is +Development of pacman is currently done in Git. The central repository is hosted by Arch Linux, although some of the developers have their own trees (ask on the above mailing lists if you are interested in finding the locations of these trees). @@ -238,7 +238,7 @@ Bugs ---- If you find bugs (which is quite likely), please email them to the pacman-dev mailing last at mailto:pacman-dev@archlinux.org[] with specific information -such as your commandline, the nature of the bug, and even the package database +such as your command-line, the nature of the bug, and even the package database if it helps. You can also post a bug to the Arch Linux bug tracker -- 1.8.4.2

1 0

[pacman-dev] [PATCH 0/2] Add color support in -Ql and option to output in tree format
by Bill Kolokithas 07 Nov '13

07 Nov '13

Hello. This is my first contribution. I Wanted to enhance -Ql output so i prepared 2 patches. The first one is quite trivial, so i don't think there will be a problem. The second one i'm not quite sure you'll want it in pacman instead of an external tool like pkgfile or pactree. I'm submitting it anyways so you can take a look. Right now you need to pass 2 (or more) -l flags in -Q(uery) to activate it. This is obviously very naive and it will break some flag combinations. This is just a proof of concept. If you like it, then a seperate flag like --tree can be added. Thanks for any feedback! Bill Kolokithas (2): Make folders appear in bold when listing files owned by a package. Add Query flag to list all files contained in a package in a tree format. src/pacman/package.c | 84 ++++++++++++++++++++++++++++++++++++++++++++++++++++ src/pacman/package.h | 1 + src/pacman/pacman.c | 2 +- src/pacman/query.c | 4 ++- 4 files changed, 89 insertions(+), 2 deletions(-) -- 1.8.4.2

3 5

[pacman-dev] Fwd: Pacman Config File (SigLevel)
by Dave Reisner 06 Nov '13

06 Nov '13

----- Forwarded message from Ingenieria Informatica <heyom.student(a)gmail.com> ----- > Date: Tue, 5 Nov 2013 22:40:18 +0100 > From: Ingenieria Informatica <heyom.student(a)gmail.com> > To: dreisner(a)archlinux.org, dan(a)archlinux.org, allan(a)archlinux.org > Subject: Pacman Config File (SigLevel) > > Hello, my name is Hamoud and I have a small question: > > According to pacman.conf: > > [options] > > # By default, pacman accepts packages signed by keys that its local keyring > # trusts (see pacman-key and its man page), as well as unsigned packages. > SigLevel = Required TrustedOnly > LocalFileSigLevel = Required TrustedOnly > > I understand from the above that signatures will be required from both packages > & databases AND If a signature is checked, it must be in the keyring and fully > trusted; marginal trust does not meet this criteria. > > Is that a valid configuration? because I only want to modify the option section > as a default control of my packages/databases. I need to harden my system as > much as possible specially with signature check (paranoid mode hehe) > > Thank you very much in advance. :) > > -- > Regards ----- End forwarded message -----

2 1

[pacman-dev] [PATCH] log important events from the backend
by Andrew Gregory 05 Nov '13

05 Nov '13

This ensures that important events will be logged and consistent regardless of the frontend. The need for global context in the event callback is also removed. The event is logged before any post_* scripts run, so this also moves the post_* script output underneath the event in the log. Fixes FS#36504 Signed-off-by: Andrew Gregory <andrew.gregory.8(a)gmail.com> --- lib/libalpm/add.c | 22 ++++++++++++++++++++++ lib/libalpm/remove.c | 5 +++++ src/pacman/callback.c | 30 ++---------------------------- 3 files changed, 29 insertions(+), 28 deletions(-) diff --git a/lib/libalpm/add.c b/lib/libalpm/add.c index ac4e36a..ee29a12 100644 --- a/lib/libalpm/add.c +++ b/lib/libalpm/add.c @@ -649,6 +649,28 @@ static int commit_single_pkg(alpm_handle_t *handle, alpm_pkg_t *newpkg, PROGRESS(handle, event, newpkg->name, 100, pkg_count, pkg_current); + switch(done) { + case ALPM_EVENT_ADD_DONE: + alpm_logaction(handle, ALPM_CALLER_PREFIX, "installed %s (%s)\n", + newpkg->name, newpkg->version); + break; + case ALPM_EVENT_DOWNGRADE_DONE: + alpm_logaction(handle, ALPM_CALLER_PREFIX, "downgraded %s (%s -> %s)\n", + newpkg->name, oldpkg->version, newpkg->version); + break; + case ALPM_EVENT_REINSTALL_DONE: + alpm_logaction(handle, ALPM_CALLER_PREFIX, "reinstalled %s (%s)\n", + newpkg->name, newpkg->version); + break; + case ALPM_EVENT_UPGRADE_DONE: + alpm_logaction(handle, ALPM_CALLER_PREFIX, "upgraded %s (%s -> %s)\n", + newpkg->name, oldpkg->version, newpkg->version); + break; + default: + /* we should never reach here */ + break; + } + /* run the post-install script if it exists */ if(alpm_pkg_has_scriptlet(newpkg) && !(trans->flags & ALPM_TRANS_FLAG_NOSCRIPTLET)) { diff --git a/lib/libalpm/remove.c b/lib/libalpm/remove.c index 7237a56..e745d89 100644 --- a/lib/libalpm/remove.c +++ b/lib/libalpm/remove.c @@ -687,6 +687,11 @@ int _alpm_remove_single_package(alpm_handle_t *handle, remove_package_files(handle, oldpkg, newpkg, targ_count, pkg_count); } + if(!newpkg) { + alpm_logaction(handle, ALPM_CALLER_PREFIX, "removed %s (%s)\n", + oldpkg->name, oldpkg->version); + } + /* run the post-remove script if it exists */ if(!newpkg && alpm_pkg_has_scriptlet(oldpkg) && !(handle->trans->flags & ALPM_TRANS_FLAG_NOSCRIPTLET)) { diff --git a/src/pacman/callback.c b/src/pacman/callback.c index 36531a2..3bdfe2c 100644 --- a/src/pacman/callback.c +++ b/src/pacman/callback.c @@ -148,8 +148,6 @@ static void fill_progress(const int bar_percent, const int disp_percent, fflush(stdout); } - - /* callback to handle messages/notifications from libalpm transactions */ void cb_event(alpm_event_t event, void *data1, void *data2) { @@ -177,10 +175,6 @@ void cb_event(alpm_event_t event, void *data1, void *data2) } break; case ALPM_EVENT_ADD_DONE: - alpm_logaction(config->handle, PACMAN_CALLER_PREFIX, - "installed %s (%s)\n", - alpm_pkg_get_name(data1), - alpm_pkg_get_version(data1)); display_optdepends(data1); break; case ALPM_EVENT_REMOVE_START: @@ -188,23 +182,12 @@ void cb_event(alpm_event_t event, void *data1, void *data2) printf(_("removing %s...\n"), alpm_pkg_get_name(data1)); } break; - case ALPM_EVENT_REMOVE_DONE: - alpm_logaction(config->handle, PACMAN_CALLER_PREFIX, - "removed %s (%s)\n", - alpm_pkg_get_name(data1), - alpm_pkg_get_version(data1)); - break; case ALPM_EVENT_UPGRADE_START: if(config->noprogressbar) { printf(_("upgrading %s...\n"), alpm_pkg_get_name(data1)); } break; case ALPM_EVENT_UPGRADE_DONE: - alpm_logaction(config->handle, PACMAN_CALLER_PREFIX, - "upgraded %s (%s -> %s)\n", - alpm_pkg_get_name(data1), - alpm_pkg_get_version(data2), - alpm_pkg_get_version(data1)); display_new_optdepends(data2, data1); break; case ALPM_EVENT_DOWNGRADE_START: @@ -213,11 +196,6 @@ void cb_event(alpm_event_t event, void *data1, void *data2) } break; case ALPM_EVENT_DOWNGRADE_DONE: - alpm_logaction(config->handle, PACMAN_CALLER_PREFIX, - "downgraded %s (%s -> %s)\n", - alpm_pkg_get_name(data1), - alpm_pkg_get_version(data2), - alpm_pkg_get_version(data1)); display_new_optdepends(data2, data1); break; case ALPM_EVENT_REINSTALL_START: @@ -225,12 +203,6 @@ void cb_event(alpm_event_t event, void *data1, void *data2) printf(_("reinstalling %s...\n"), alpm_pkg_get_name(data1)); } break; - case ALPM_EVENT_REINSTALL_DONE: - alpm_logaction(config->handle, PACMAN_CALLER_PREFIX, - "reinstalled %s (%s)\n", - alpm_pkg_get_name(data1), - alpm_pkg_get_version(data1)); - break; case ALPM_EVENT_INTEGRITY_START: if(config->noprogressbar) { printf(_("checking package integrity...\n")); @@ -286,6 +258,8 @@ void cb_event(alpm_event_t event, void *data1, void *data2) } break; /* all the simple done events, with fallthrough for each */ + case ALPM_EVENT_REINSTALL_DONE: + case ALPM_EVENT_REMOVE_DONE: case ALPM_EVENT_FILECONFLICTS_DONE: case ALPM_EVENT_CHECKDEPS_DONE: case ALPM_EVENT_RESOLVEDEPS_DONE: -- 1.8.4.2

1 0

[pacman-dev] [RFC][PATCH 0/3] Add source package information to repo database
by Allan McRae 05 Nov '13

05 Nov '13

AKA - the I hate ABS patchset... This patchset adds source package information to the sync dbs. Once this is given the all clear, it will be followed by patches reading the information into libalpm and adding a "pacman -B" that will download the source package and extract to /var/build/pacman (or somewhere...). This would be a complete ABS replacement on Arch. Queries: 1) I add a separate file to the db and reuse the variable names from the desc file. This means some probably duplication in the sync db reading code. Or should I add the info to the desc file and give it different names? 2) I add a -S/--source flag to add the source file description. This is similar to --delta and --files. I toyed with the idea of including the source information by default. Opinions on this? Allan McRae (3): makepkg: sign source packages with --sign Always supply base name and version info in .PKGFILE if needed repo-add: add source package information to sync db lib/libalpm/be_package.c | 2 ++ lib/libalpm/be_sync.c | 2 ++ scripts/makepkg.sh.in | 30 +++++++++++++------ scripts/repo-add.sh.in | 75 ++++++++++++++++++++++++++++++++++++++++++++++-- 4 files changed, 99 insertions(+), 10 deletions(-) -- 1.8.4.2

2 6

[pacman-dev] [PATCH] pacsort: only look at versions in file mode
by Dave Reisner 05 Nov '13

05 Nov '13

alpm_pkg_vercmp rightfully considers foo-1-1 to be newer than foo-1:1-1 because of the leading characters. So, in file mode, pacsort must necessarily ignore the package name when performing comparisons. Create a struct, for convenience, which overlays pointers and lengths to the name and version within a filename. Using this, we can avoid some amount of heap allocation, and do better at comparing versions for similar package names. Fixes FS#37631. Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> --- src/util/pacsort.c | 69 +++++++++++++++++++++++++++++++++++++++++------- test/util/pacsorttest.sh | 11 +++++++- 2 files changed, 69 insertions(+), 11 deletions(-) diff --git a/src/util/pacsort.c b/src/util/pacsort.c index 687e558..cbee316 100644 --- a/src/util/pacsort.c +++ b/src/util/pacsort.c @@ -28,6 +28,15 @@ #define DELIM ' ' +#ifndef MIN +#define MIN(a, b) \ + __extension__({ \ + __typeof__(a) _a = (a); \ + __typeof__(b) _b = (b); \ + _a < _b ? _a : _b; \ + }) +#endif + struct buffer_t { char *mem; size_t len; @@ -40,6 +49,14 @@ struct list_t { size_t maxcount; }; +struct pkgmeta_t { + const char *pkgname; + size_t pkgname_len; + + const char *pkgver; + size_t pkgver_len; +}; + static struct options_t { int order; int sortkey; @@ -258,6 +275,30 @@ static const char *nth_column(const char *string) return prev; } +static void extract_pkgmeta(const char *path, struct pkgmeta_t *meta) { + const char *pkgver_end; + + memset(meta, 0, sizeof(struct pkgmeta_t)); + + meta->pkgname = basename(path); + pkgver_end = strrchr(meta->pkgname, '-'); + + /* read backwards through pkgrel */ + for(meta->pkgver = pkgver_end - 1; + meta->pkgver > meta->pkgname && *meta->pkgver != '-'; + --meta->pkgver) + ; + /* read backwards through pkgver */ + for(--meta->pkgver; + meta->pkgver > meta->pkgname && *meta->pkgver != '-'; + --meta->pkgver) + ; + ++meta->pkgver; + + meta->pkgname_len = meta->pkgver - meta->pkgname - 1; + meta->pkgver_len = pkgver_end - meta->pkgver; +} + static int vercmp(const void *p1, const void *p2) { const char *name1, *name2; @@ -280,17 +321,25 @@ static int vercmp(const void *p1, const void *p2) if(opts.filemode) { if(fnmatch("*-*.pkg.tar.?z", name1, 0) == 0 && fnmatch("*-*.pkg.tar.?z", name2, 0) == 0) { - const char *start, *end; - - start = strrchr(name1, '/'); - start = start ? start + 1 : name1; - end = strrchr(name1, '-'); - fn1 = strndup(start, end - start); + struct pkgmeta_t pkgmeta1, pkgmeta2; + int namecmp; + + /* Extract name and version */ + extract_pkgmeta(name1, &pkgmeta1); + extract_pkgmeta(name2, &pkgmeta2); + + /* If the package names aren't the same, there's no sense in comparing + * the versions. Short circuit and return a comparison based on the + * pkgnames. This probably could be cleaner, but I'm favoring clarity + * over cleverness. */ + namecmp = memcmp(pkgmeta1.pkgname, pkgmeta2.pkgname, + MIN(pkgmeta1.pkgname_len, pkgmeta2.pkgname_len)); + if(pkgmeta1.pkgname_len != pkgmeta2.pkgname_len || namecmp != 0) { + return opts.order * namecmp; + } - start = strrchr(name2, '/'); - start = start ? start + 1 : name2; - end = strrchr(name2, '-'); - fn2 = strndup(start, end - start); + fn1 = strndup(pkgmeta1.pkgver, pkgmeta1.pkgver_len); + fn2 = strndup(pkgmeta2.pkgver, pkgmeta2.pkgver_len); name1 = fn1; name2 = fn2; diff --git a/test/util/pacsorttest.sh b/test/util/pacsorttest.sh index ac16c45..8382603 100755 --- a/test/util/pacsorttest.sh +++ b/test/util/pacsorttest.sh @@ -21,7 +21,7 @@ # default binary if one was not specified as $1 bin=${1:-${PMTEST_UTIL_DIR}pacsort} # holds counts of tests -total=23 +total=26 run=0 failure=0 @@ -89,6 +89,15 @@ runtest $in $in "filename sort with uneven leading path components" "--files" in="firefox-18.0-2-i686.pkg.tar.xz\nfirefox-18.0.1-1-x86_64.pkg.tar.gz\n" runtest $in $in "filename sort with different extensions" "--files" +in="/packages/dialog-1.2_20131001-1-x86_64.pkg.tar.xz\n/packages/dialog-1:1.2_20130928-1-x86_64.pkg.tar.xz\n" +runtest $in $in "filename sort with epoch" "--files" + +in="/packages/dia-log-1:1.2_20130928-1-x86_64.pkg.tar.xz\n/packages/dialog-1.2_20131001-1-x86_64.pkg.tar.xz\n" +runtest $in $in "filename sort with epoch" "--files" + +in="/packages/dialoggg-1:1.2_20130928-1-x86_64.pkg.tar.xz\n/packages/dialog-1.2_20131001-1-x86_64.pkg.tar.xz\n" +runtest $in $in "filename sort with epoch" "--files" + # generate some long input/expected for the next few tests declare normal reverse names_normal names_reverse for ((i=1; i<600; i++)); do -- 1.8.4.2

3 2

[pacman-dev] [PATCH] Makefile.am: fix typo in LOG_DRIVER variable
by Andrew Gregory 05 Nov '13

05 Nov '13

Self-executing tests were not being run through the tap log driver. This caused `make check` to ignore discrepancies between the expected number of tests and the actual number of tests. Also, fix some uncommented output from test scripts that could confuse TAP parsers. Signed-off-by: Andrew Gregory <andrew.gregory.8(a)gmail.com> --- As a side effect, each test in the scripts is counted individually by make check, giving us a more impressive test count of 446. Makefile.am | 2 +- test/scripts/human_to_size_test.sh | 2 +- test/scripts/parseopts_test.sh | 2 +- test/util/pacsorttest.sh | 4 ++-- test/util/vercmptest.sh | 2 +- 5 files changed, 6 insertions(+), 6 deletions(-) diff --git a/Makefile.am b/Makefile.am index 0ba39ca..4d5adae 100644 --- a/Makefile.am +++ b/Makefile.am @@ -36,7 +36,7 @@ TEST_EXTENSIONS = .py AM_TESTS_ENVIRONMENT = \ PMTEST_UTIL_DIR=$(top_builddir)/src/util/; export PMTEST_UTIL_DIR; \ PMTEST_SCRIPTLIB_DIR=$(top_srcdir)/scripts/library/; export PMTEST_SCRIPTLIB_DIR; -TEST_LOG_DRIVER = env AM_TAP_AWK='$(AWK)' $(SHELL) \ +LOG_DRIVER = env AM_TAP_AWK='$(AWK)' $(SHELL) \ $(top_srcdir)/build-aux/tap-driver.sh PY_LOG_DRIVER = env AM_TAP_AWK='$(AWK)' $(SHELL) \ $(top_srcdir)/build-aux/tap-driver.sh diff --git a/test/scripts/human_to_size_test.sh b/test/scripts/human_to_size_test.sh index 6306137..a467f27 100755 --- a/test/scripts/human_to_size_test.sh +++ b/test/scripts/human_to_size_test.sh @@ -10,7 +10,7 @@ if [[ -z $lib || ! -f $lib ]]; then fi . "$lib" -if ! type -t human_to_size >/dev/null; then +if ! type -t human_to_size &>/dev/null; then printf 'Bail out! human_to_size function not found\n' exit 1 fi diff --git a/test/scripts/parseopts_test.sh b/test/scripts/parseopts_test.sh index 5ff4bc5..00bdbd8 100755 --- a/test/scripts/parseopts_test.sh +++ b/test/scripts/parseopts_test.sh @@ -10,7 +10,7 @@ if [[ -z $lib || ! -f $lib ]]; then fi . "$lib" -if ! type -t parseopts >/dev/null; then +if ! type -t parseopts &>/dev/null; then printf 'Bail out! parseopts function not found\n' exit 1 fi diff --git a/test/util/pacsorttest.sh b/test/util/pacsorttest.sh index ac16c45..ef24a76 100755 --- a/test/util/pacsorttest.sh +++ b/test/util/pacsorttest.sh @@ -25,7 +25,7 @@ total=23 run=0 failure=0 -if ! type -p "$bin"; then +if ! type -p "$bin" &>/dev/null; then echo "Bail out! pacsort binary ($bin) could not be located" exit 1 fi @@ -47,7 +47,7 @@ runtest() { fi } -echo "Running pacsort tests..." +echo "# Running pacsort tests..." echo "1..$total" diff --git a/test/util/vercmptest.sh b/test/util/vercmptest.sh index a7fd851..2d74b57 100755 --- a/test/util/vercmptest.sh +++ b/test/util/vercmptest.sh @@ -25,7 +25,7 @@ run=0 failure=0 # use first arg as our binary if specified -if ! type -p "$bin"; then +if ! type -p "$bin" &>/dev/null; then echo "Bail out! vercmp binary ($bin) could not be located" exit 1 fi -- 1.8.4.2

1 0

[pacman-dev] Repo database(s) layout
by Allan McRae 05 Nov '13

05 Nov '13

Hi, We currently have a .db and .files databases, with .files being a superset of .db. An idea was formed on IRC to completely separate these. I.e. .db stays as it is and .files only includes the file lists. We would then add .source to include the source package information. I would set repo-add to automatically create all these files. We would then add something like "-S --refresh-files" and "-S --refresh-source" to download those files as a one off, printing a warning when using them if they are out of date compared to the repo. Another option is to use Usage as a flag for when to download them, but refreshing all those every update seems excessive. This would also allow us to have some basic pkgfile functionality in pacman (-So). So, there much to work out, but does the general idea sound good to people? Allan

2 2

[pacman-dev] [PATCH] Ignore makedepend and checkdepend entries when parsing .PKGINFO file
by Allan McRae 01 Nov '13

01 Nov '13

Signed-off-by: Allan McRae <allan(a)archlinux.org> --- lib/libalpm/be_package.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/lib/libalpm/be_package.c b/lib/libalpm/be_package.c index 3f577ba..4e814f8 100644 --- a/lib/libalpm/be_package.c +++ b/lib/libalpm/be_package.c @@ -218,6 +218,10 @@ static int parse_descfile(alpm_handle_t *handle, struct archive *a, alpm_pkg_t * } else if(strcmp(key, "optdepend") == 0) { alpm_depend_t *optdep = _alpm_splitdep(ptr); newpkg->optdepends = alpm_list_add(newpkg->optdepends, optdep); + } else if(strcmp(key, "makedepend") == 0) { + /* not used atm */ + } else if(strcmp(key, "checkdepend") == 0) { + /* not used atm */ } else if(strcmp(key, "conflict") == 0) { alpm_depend_t *conflict = _alpm_splitdep(ptr); newpkg->conflicts = alpm_list_add(newpkg->conflicts, conflict); -- 1.8.4.2

1 0

[pacman-dev] [PATCH] Return zero from extract_file with local source file
by Allan McRae 01 Nov '13

01 Nov '13

The call to bsdtar to check if a file needs extracted returns 1 if it does not. We then propegate this return value accientally which can cause makepkg to report an error later on. Explicitly return 0 in this case. Signed-off-by: Allan McRae <allan(a)archlinux.org> --- scripts/makepkg.sh.in | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/scripts/makepkg.sh.in b/scripts/makepkg.sh.in index 0fc71c0..401d681 100644 --- a/scripts/makepkg.sh.in +++ b/scripts/makepkg.sh.in @@ -419,7 +419,7 @@ extract_file() { if bsdtar -tf "$file" -q '*' &>/dev/null; then cmd="bsdtar" else - return + return 0 fi ;; esac -- 1.8.4.2

1 1