- Pacman-dev - lists.archlinux.org

[pacman-dev] [PATCH] makepkg: allow skipping intergrity checks whem making source package
by Allan McRae 11 Oct '09

11 Oct '09

Extends the use of the --skipinteg option to creating a source package. Fixes FS#15984. Signed-off-by: Allan McRae <allan(a)archlinux.org> --- scripts/makepkg.sh.in | 10 +++++++--- 1 files changed, 7 insertions(+), 3 deletions(-) diff --git a/scripts/makepkg.sh.in b/scripts/makepkg.sh.in index 88487da..d85f858 100644 --- a/scripts/makepkg.sh.in +++ b/scripts/makepkg.sh.in @@ -1038,9 +1038,13 @@ create_srcpackage() { # Get back to our src directory so we can begin with sources. mkdir -p "$srcdir" cd "$srcdir" - download_sources - # We can only check checksums if we have all files. - check_checksums + if [ $SKIPINTEG -eq 0 ]; then + # We can only check checksums if we have all files. + download_sources + check_checksums + else + warning "$(gettext "Skipping integrity checks.")" + fi cd "$startdir" msg "$(gettext "Creating source package...")" -- 1.6.4.4

1 1

[pacman-dev] [PATCH 1/5] Rework the alpm_unpack functions
by Xavier Chantry 11 Oct '09

11 Oct '09

Add support to extract a list of entries Signed-off-by: Xavier Chantry <shiningxc(a)gmail.com> --- lib/libalpm/trans.c | 2 +- lib/libalpm/util.c | 55 ++++++++++++++++++++++++++++++++++++++++---------- lib/libalpm/util.h | 5 +++- 3 files changed, 49 insertions(+), 13 deletions(-) diff --git a/lib/libalpm/trans.c b/lib/libalpm/trans.c index c99f596..6f5216e 100644 --- a/lib/libalpm/trans.c +++ b/lib/libalpm/trans.c @@ -368,7 +368,7 @@ int _alpm_runscriptlet(const char *root, const char *installfn, /* either extract or copy the scriptlet */ snprintf(scriptfn, PATH_MAX, "%s/.INSTALL", tmpdir); if(!strcmp(script, "pre_upgrade") || !strcmp(script, "pre_install")) { - if(_alpm_unpack(installfn, tmpdir, ".INSTALL")) { + if(_alpm_unpack_single(installfn, tmpdir, ".INSTALL")) { retval = 1; } } else { diff --git a/lib/libalpm/util.c b/lib/libalpm/util.c index 381b116..5785497 100644 --- a/lib/libalpm/util.c +++ b/lib/libalpm/util.c @@ -241,14 +241,36 @@ int _alpm_lckrm() /* Compression functions */ /** - * @brief Unpack a specific file or all files in an archive. + * @brief Unpack a specific file in an archive. * * @param archive the archive to unpack * @param prefix where to extract the files - * @param fn a file within the archive to unpack or NULL for all + * @param fn a file within the archive to unpack * @return 0 on success, 1 on failure */ -int _alpm_unpack(const char *archive, const char *prefix, const char *fn) +int _alpm_unpack_single(const char *archive, const char *prefix, const char *fn) +{ + alpm_list_t *list = NULL; + int ret = 0; + if(fn == NULL) { + return(1); + } + list = alpm_list_add(list, (void *)fn); + ret = _alpm_unpack(archive, prefix, &list); + alpm_list_free(list); + return(ret); +} + +/** + * @brief Unpack a list of files in an archive. + * + * @param archive the archive to unpack + * @param prefix where to extract the files + * @param list a pointer to a list of files within the archive to unpack or + * NULL for all + * @return 0 on success, 1 on failure + */ +int _alpm_unpack(const char *archive, const char *prefix, alpm_list_t **list) { int ret = 0; mode_t oldmask; @@ -259,6 +281,10 @@ int _alpm_unpack(const char *archive, const char *prefix, const char *fn) ALPM_LOG_FUNC; + if(list && *list == NULL) { + return(0); + } + if((_archive = archive_read_new()) == NULL) RET_ERR(PM_ERR_LIBARCHIVE, 1); @@ -301,14 +327,18 @@ int _alpm_unpack(const char *archive, const char *prefix, const char *fn) archive_entry_set_perm(entry, 0755); } - /* If a specific file was requested skip entries that don't match. */ - if (fn && strcmp(fn, entryname)) { - _alpm_log(PM_LOG_DEBUG, "skipping: %s\n", entryname); - if (archive_read_data_skip(_archive) != ARCHIVE_OK) { - ret = 1; - goto cleanup; + /* If specific files were requested, skip entries that don't match. */ + if(list && *list != NULL) { + char *data = NULL; + *list = alpm_list_remove_str(*list, entryname, &data); + if(data == NULL) { + _alpm_log(PM_LOG_DEBUG, "skipping: %s\n", entryname); + if (archive_read_data_skip(_archive) != ARCHIVE_OK) { + ret = 1; + goto cleanup; + } + continue; } - continue; } /* Extract the archive entry. */ @@ -324,12 +354,15 @@ int _alpm_unpack(const char *archive, const char *prefix, const char *fn) goto cleanup; } - if(fn) { + if(list && *list == NULL) { break; } } cleanup: + if(list && *list != NULL) { + _alpm_log(PM_LOG_DEBUG, "not all entries were extracted\n"); + } umask(oldmask); archive_read_finish(_archive); if(restore_cwd) { diff --git a/lib/libalpm/util.h b/lib/libalpm/util.h index ccf169d..58a8607 100644 --- a/lib/libalpm/util.h +++ b/lib/libalpm/util.h @@ -26,6 +26,8 @@ #include "config.h" +#include "alpm_list.h" + #include <stdio.h> #include <string.h> #include <stdarg.h> @@ -62,7 +64,8 @@ int _alpm_copyfile(const char *src, const char *dest); char *_alpm_strtrim(char *str); int _alpm_lckmk(); int _alpm_lckrm(); -int _alpm_unpack(const char *archive, const char *prefix, const char *fn); +int _alpm_unpack_single(const char *archive, const char *prefix, const char *fn); +int _alpm_unpack(const char *archive, const char *prefix, alpm_list_t **list); int _alpm_rmrf(const char *path); int _alpm_logaction(unsigned short usesyslog, FILE *f, const char *fmt, va_list args); int _alpm_run_chroot(const char *root, const char *cmd); -- 1.6.4.4

2 5

[pacman-dev] [PATCH] [RFC] Only extract new db entries
by Allan McRae 09 Oct '09

09 Oct '09

This is an incomplete patch that I need some help finishing. The idea is to only extract folders for new packages into the package database and then clean up the old directories. This is essentially implementing Xyne's "rebase" script within pacman. So far I have got: 1. Only remove all db entries if using "-Syy" 2. Generate list of directories in db 3. ??? 4. Clean up directories not found in new db. Step 3 needs to check if an directory in the sync db archive is in the current database. If it is not, then extract its files. If it is, then remove it from the list of directories that will be removed in step 4. This is where I am confused on how best to proceed. The archive will be open to read a list of files, so calling _aplm_unpack for a single file would seem a waste. I am not sure how to continue. Signed-off-by: Allan McRae <allan(a)archlinux.org> --- lib/libalpm/be_files.c | 60 ++++++++++++++++++++++++++++++++++++++++++----- 1 files changed, 53 insertions(+), 7 deletions(-) diff --git a/lib/libalpm/be_files.c b/lib/libalpm/be_files.c index 21533ef..5e5a5fa 100644 --- a/lib/libalpm/be_files.c +++ b/lib/libalpm/be_files.c @@ -181,6 +181,9 @@ int SYMEXPORT alpm_db_update(int force, pmdb_t *db) time_t newmtime = 0, lastupdate = 0; const char *dbpath; size_t len; + DIR *dbdir; + struct dirent *ent = NULL; + alpm_list_t *dbdirlist; int ret; @@ -229,36 +232,79 @@ int SYMEXPORT alpm_db_update(int force, pmdb_t *db) return(-1); } else { const char *syncdbpath = _alpm_db_path(db); - /* remove the old dir */ - if(_alpm_rmrf(syncdbpath) != 0) { - _alpm_log(PM_LOG_ERROR, _("could not remove database %s\n"), db->treename); - RET_ERR(PM_ERR_DB_REMOVE, -1); + + /* remove the old dir if forcing update */ + if(force) { + if(_alpm_rmrf(syncdbpath) != 0) { + _alpm_log(PM_LOG_ERROR, _("could not remove database %s\n"), db->treename); + RET_ERR(PM_ERR_DB_REMOVE, -1); + } } - /* Cache needs to be rebuilt */ - _alpm_db_free_pkgcache(db); + /* create list of directories in db */ + dbdir = opendir(syncdbpath); + if (dbdir != NULL) { + while((ent = readdir(dbdir)) != NULL) { + const char *name = ent->d_name + if(strcmp(name, ".") == 0 || strcmp(name, "..") == 0) { + continue; + } + + /* stat the entry, make sure it's a directory */ + snprintf(path, PATH_MAX, "%s%s", dbpath, name); + if(stat(path, &sbuf) != 0 || !S_ISDIR(sbuf.st_mode)) { + continue; + } + + dbdirlist = _alpm_list_add(dbdirlist, name) + } + } + close(dbdir); /* form the path to the db location */ len = strlen(dbpath) + strlen(db->treename) + strlen(DBEXT) + 1; MALLOC(dbfilepath, len, RET_ERR(PM_ERR_MEMORY, -1)); sprintf(dbfilepath, "%s%s" DBEXT, dbpath, db->treename); - /* uncompress the sync database */ + /* TODO - extract new entries from the sync database */ + /* Remove old entries from dbdirlist as we go */ checkdbdir(db); + +/* old code to be replaced ret = _alpm_unpack(dbfilepath, syncdbpath, NULL); if(ret) { free(dbfilepath); RET_ERR(PM_ERR_SYSTEM, -1); } +*/ + unlink(dbfilepath); free(dbfilepath); + /* remove old directories from dbdir */ + for (i = dbdirlist; i; i = i->next) { + const char *name = i->data; + len = strlen(dbpath) + strlen(db->treename) + strlen(DBEXT) + 1; + MALLOC(olddbdir, len, RET_ERR(PM_ERR_MEMORY, -1)); + snprintf(olddbdir, PATH_MAX, "%s/%s", syncdbpath, name); + if(_alpm_rmrf(olddbdir) != 0) { + _alpm_log(PM_LOG_ERROR, _("could not remove database directory %s\n"), olddbdir); + free(olddbdir); + RET_ERR(PM_ERR_DB_REMOVE, -1); + } + free(olddbdir); + } + FREELIST(dbdirlist); + /* if we have a new mtime, set the DB last update value */ if(newmtime) { _alpm_log(PM_LOG_DEBUG, "sync: new mtime for %s: %ju\n", db->treename, (uintmax_t)newmtime); setlastupdate(db, newmtime); } + + /* Cache needs to be rebuilt */ + _alpm_db_free_pkgcache(db); } return(0); -- 1.6.4.4

1 0

[pacman-dev] [PATCH] Show 'Required By' in -Si output
by Dan McGee 09 Oct '09

09 Oct '09

Just as we do in -Qi, we can compute required by information for sync database packages. The behavior seems sane; for a given package, the -Si required by will show all packages in *any* sync database that require it. Implements FS#16244. Signed-off-by: Dan McGee <dan(a)archlinux.org> --- lib/libalpm/package.c | 46 ++++++++++++++++++++++++++++++++++++---------- src/pacman/package.c | 13 ++++--------- 2 files changed, 40 insertions(+), 19 deletions(-) diff --git a/lib/libalpm/package.c b/lib/libalpm/package.c index de17166..b0b6480 100644 --- a/lib/libalpm/package.c +++ b/lib/libalpm/package.c @@ -556,6 +556,21 @@ unsigned short SYMEXPORT alpm_pkg_has_scriptlet(pmpkg_t *pkg) return pkg->scriptlet; } +static void find_requiredby(pmpkg_t *pkg, pmdb_t *db, alpm_list_t **reqs) +{ + const alpm_list_t *i; + for(i = _alpm_db_get_pkgcache(db); i; i = i->next) { + if(!i->data) { + continue; + } + pmpkg_t *cachepkg = i->data; + if(_alpm_dep_edge(cachepkg, pkg)) { + const char *cachepkgname = alpm_pkg_get_name(cachepkg); + *reqs = alpm_list_add(*reqs, strdup(cachepkgname)); + } + } +} + /** * @brief Compute the packages requiring a given package. * @param pkg a package @@ -565,18 +580,29 @@ alpm_list_t SYMEXPORT *alpm_pkg_compute_requiredby(pmpkg_t *pkg) { const alpm_list_t *i; alpm_list_t *reqs = NULL; + pmdb_t *db; - pmdb_t *localdb = alpm_option_get_localdb(); - for(i = _alpm_db_get_pkgcache(localdb); i; i = i->next) { - if(!i->data) { - continue; - } - pmpkg_t *cachepkg = i->data; - if(_alpm_dep_edge(cachepkg, pkg)) { - const char *cachepkgname = alpm_pkg_get_name(cachepkg); - reqs = alpm_list_add(reqs, strdup(cachepkgname)); + if(pkg->origin == PKG_FROM_FILE) { + /* The sane option; search locally for things that require this. */ + db = alpm_option_get_localdb(); + fprintf(stderr, "db name: %s\n", db->treename); + find_requiredby(pkg, db, &reqs); + } else { + /* We have a DB package. if it is a local package, then we should + * only search the local DB; else search all known sync databases. */ + db = pkg->origin_data.db; + if(db->is_local) { + fprintf(stderr, "db name: %s\n", db->treename); + find_requiredby(pkg, db, &reqs); + } else { + for(i = handle->dbs_sync; i; i = i->next) { + db = i->data; + fprintf(stderr, "db name: %s\n", db->treename); + find_requiredby(pkg, db, &reqs); + } } } + return(reqs); } diff --git a/src/pacman/package.c b/src/pacman/package.c index 3b14516..e7e2552 100644 --- a/src/pacman/package.c +++ b/src/pacman/package.c @@ -83,10 +83,8 @@ void dump_pkg_full(pmpkg_t *pkg, int level) depstrings = alpm_list_add(depstrings, alpm_dep_compute_string(dep)); } - if(level>0) { - /* compute this here so we don't get a pause in the middle of output */ - requiredby = alpm_pkg_compute_requiredby(pkg); - } + /* compute this here so we don't get a pause in the middle of output */ + requiredby = alpm_pkg_compute_requiredby(pkg); /* actual output */ string_display(_("Name :"), alpm_pkg_get_name(pkg)); @@ -97,11 +95,7 @@ void dump_pkg_full(pmpkg_t *pkg, int level) list_display(_("Provides :"), alpm_pkg_get_provides(pkg)); list_display(_("Depends On :"), depstrings); list_display_linebreak(_("Optional Deps :"), alpm_pkg_get_optdepends(pkg)); - /* Only applicable if installed */ - if(level > 0) { - list_display(_("Required By :"), requiredby); - FREELIST(requiredby); - } + list_display(_("Required By :"), requiredby); list_display(_("Conflicts With :"), alpm_pkg_get_conflicts(pkg)); list_display(_("Replaces :"), alpm_pkg_get_replaces(pkg)); if(level < 0) { @@ -142,6 +136,7 @@ void dump_pkg_full(pmpkg_t *pkg, int level) printf("\n"); FREELIST(depstrings); + FREELIST(requiredby); } /* Display the content of a sync package -- 1.6.4.4

8 11

[pacman-dev] 3.3.2 release soon
by Xavier 08 Oct '09

08 Oct '09

Looks like we are going to need a new bugfix release sooner than expected. Just when 3.3.1 was released, two issues popped up : 1) the repo-add / CSIZE / any / symlink issue : http://code.toofishes.net/cgit/dan/pacman.git/commit/?h=maint&id=20aa17c276… 2) the infinite download issue : http://bugs.archlinux.org/task/16359 That's also an occasion for translators to do a quick update if needed (but there are no modified/new strings). Especially Swedish and Chinese since these two were not fully translated for 3.3.1

4 15

[pacman-dev] [GIT] The official pacman repository annotated tag, v3.3.2, created. v3.3.2
by dan＠archlinux.org 07 Oct '09

07 Oct '09

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 annotated tag, v3.3.2 has been created at d8aab797afed17f1af2450d1e5c3cbcf8d7135e6 (tag) tagging 9282a4a9094ddeae8d0bef0f575e45c574591c4b (commit) replaces v3.3.1 tagged by Dan McGee on Wed Oct 7 18:15:48 2009 -0500 - Log ----------------------------------------------------------------- Too-quick maint release -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (GNU/Linux) iEYEABECAAYFAkrNITkACgkQXC5GoPU6du1iagCfU8lTkXHmLh6BbRI5MUwhKJcU fWMAnRN2lEjeTldSYNJDNG1J/cV62Ssi =QDEv -----END PGP SIGNATURE----- Christian Larsson (1): Update Swedish translation Dan McGee (5): repo-add: make sure we get size of target file repo-remove: create empty DB when all packages are removed doc: ensure SyncFirst behavior with flags is clear doc: Add HACKING to html target 3.2.2 release changes Xavier Chantry (5): dload : fix infinite download (big type mistake) callback.c : fallback to normal download with bogus size Update all pot and po files for 3.3.2 release Update translations New Greek translation 甘露(Lu.Gan) (1): Update chinese translation ----------------------------------------------------------------------- hooks/post-receive -- The official pacman repository

1 0

[pacman-dev] ILoveCandy misc option in pacman.conf
by Laszlo Papp 06 Oct '09

06 Oct '09

Hello! What about ILoveCandy option in pacman.conf ? It occurs in the source code, in the config parser function, but there isn't any word about it in the pacman.conf manual and it doesn't occur in the default pacman.conf file in disabled/commented line, like the other Misc options do. Will it be supported in the future so that to be worth to do the relevant manual part and pacman.conf part or not ? Best Regards, Laszlo Papp

3 2

Re: [pacman-dev] pacman 3.3.2
by Matthias Gorissen 04 Oct '09

04 Oct '09

German translation: "Keine Pakete mehr vorhanden, erstelle leere Datenbank." Am Sunday 04 October 2009 16:59:09 schrieben Sie: > Hello translation team :) > > I just realized we actually have one string change in repo-add. > > msgid "No packages remain, creating empty database." > > If you want this message translated in 3.3.2, please send me the > translation inline (as an answer to that mail). Otherwise no big deal, > its just a repo-add (actually repo-remove) which will be very rarely > used as you can guess.

1 0

[pacman-dev] two patches for makepkg
by huntxu 02 Oct '09

02 Oct '09

1. let makepkg work with icc this patch is inspired by the gentoo wiki page http://en.gentoo-wiki.com/wiki/ICC First of all, you should install icc of course, see http://aur.archlinux.org/packages.php?ID=2252 Add your icc CFLAGS/CXXFLAGS($ICCCFLAGS & $ICCCXXFLAGS) to /etc/makepkg.conf. Add "icc" to the BUILDENV array in /etc/makepkg.conf Add packages' names (one per line) you want to compiled with icc to /etc/package.icc, if a package is not listed in this file, it will be compiled with gcc. Add per-package CFLAGS for packages(package's name followed by CFLAGS per line) you want to compiled with icc to /etc/package.icc-cflags. A few packages benefit greatly from the -O3 flag like python and sqlite according to the gentoo wiki. 2. let makepkg install custom license automatically When making packages whose licenses are "custom", we use PKGBUILD to install the license file manually. Since it is a common step for those packages, why not let makepkg do that automatically? Just specify $licensefile in your PKGBUILD(like $licensefile=LICENSE) and ensure the license file is in $startdir, it then will be automatically install to $pkgdir/usr/share/licenses/$pkgname/$licensefile when packaging. -- best regards, hunt

6 6

[pacman-dev] pkgbuild parser in python
by Xavier 01 Oct '09

01 Oct '09

On Thu, Oct 1, 2009 at 4:22 PM, Sebastian Nowicki <sebnow(a)gmail.com> wrote: > > On Oct 1, 2009, at 2:52 PM, Allan McRae wrote: > >> Hi, >> >> This is one part of the makepkg test suite I am working on. It is fairly >> a simple class that takes a pacman package filename and does some parsing. >> Currently that involves getting the file list and the info from the >> .PKGINFO file. >> >> I am still fairly new to python so I am looking for comments on how this >> could be improved. >> One idea I have had is to not initialize all the fields in the pkginfo >> dict and add a test if an array exists before appending. It would make the >> code tidier but that would mean the need to test if the field exists when >> comparing it later. > > I didn't really look at the script, but I made a similar parser [1] earlier, > initially for AUR2 purposes. Looking over it, it seems I don't include the > file list during parsing, but that can probably be easily added in. Perhaps > it can be expanded upon, instead of creating a new one. If not it should at > least help. > > [1] http://github.com/sebnow/parched/blob/master/parched.py > > So this is both a pkgbuild and pkginfo parser ? For the AUR, you just need pkgbuild parser, right ? And this is much more complex than pkginfo parser. Actually that's why I hate pkgbuild :) There was another pkgbuild parser in python , written by stonecrest, and used there : http://projects.archlinux.org/?p=dbscripts.git;a=blob;f=cron-jobs/check_arc… But I switched to using bash, because this is just much simpler and works much better (well this has changed with split pkgbuilds though :P). Anyway, go Xyne :P http://xyne.archlinux.ca/ideas/pkgmeta

3 2