August 2008 - Pacman-dev - lists.archlinux.org

[pacman-dev] [BUG] alpm_list_remove need an initalized pointer for void **data
by solsTiCe d'Hiver 08 Aug '08

08 Aug '08

hi. i am playing with alpm and db4 to make a little program for me. just for fun. and in this new release of libalpm (along pacman 3.2.0) i have found that alpm_list_remove behave differently than before. so the last parameter (void **data) need to be initialized to avoid a seg fault. it happens in alpm_db_unregister if you unregister a sync db if you run this, you will see it happen #include <stdio.h> #include <stdlib.h> #include <alpm.h> int main(void) { pmdb_t *db; pmpkg_t *pkg; const char *s; alpm_initialize(); alpm_option_set_root("/"); alpm_option_set_dbpath("/var/lib/pacman"); alpm_option_add_cachedir("/var/cache/pacman/pkg"); alpm_option_set_logfile("/dev/stdout"); db = alpm_db_register_sync("extra"); puts("alpm_db_unregister(db)"); alpm_db_unregister(db); puts("alpm_release()"); alpm_release(); exit(EXIT_SUCCESS); } i wonder how it has not generated more bug in libalpm. so i check all uninitalised pointer in libalpm before alpm_list_remove and i found 3 occurences see the patch diff --git a/lib/libalpm/cache.c b/lib/libalpm/cache.c index 032cc97..b7681db 100644 --- a/lib/libalpm/cache.c +++ b/lib/libalpm/cache.c @@ -135,7 +135,7 @@ int _alpm_db_add_pkgincache(pmdb_t *db, pmpkg_t *pkg) int _alpm_db_remove_pkgfromcache(pmdb_t *db, pmpkg_t *pkg) { - void *vdata; + void *vdata=NULL; pmpkg_t *data; ALPM_LOG_FUNC; diff --git a/lib/libalpm/db.c b/lib/libalpm/db.c index d9a3931..c0d322c 100644 --- a/lib/libalpm/db.c +++ b/lib/libalpm/db.c @@ -145,7 +145,7 @@ int SYMEXPORT alpm_db_unregister(pmdb_t *db) * databases by walking through the list returned by * alpm_option_get_syncdbs, because the db is removed from that list here. */ - void *data; + void *data=NULL; handle->dbs_sync = alpm_list_remove(handle->dbs_sync, db, _alpm_db_cmp, &data); if(data) { diff --git a/lib/libalpm/remove.c b/lib/libalpm/remove.c index 864fafa..6ea16c0 100644 --- a/lib/libalpm/remove.c +++ b/lib/libalpm/remove.c @@ -118,7 +118,7 @@ static void remove_prepare_keep_needed(pmtrans_t *trans, pmdb_t *db, alpm_list_t *i; for(i = lp; i; i = i->next) { pmdepmissing_t *miss = (pmdepmissing_t *)i->data; - void *vpkg; + void *vpkg=NULL; pmpkg_t *pkg = _alpm_pkg_find(trans->packages, miss->causingpkg); if(pkg == NULL) { continue;

3 13

Re: [pacman-dev] [arch-general] rotate pacman log?
by Aaron Griffin 08 Aug '08

08 Aug '08

On Thu, Aug 7, 2008 at 4:05 PM, Dan McGee <dpmcgee(a)gmail.com> wrote: > On Thu, Aug 7, 2008 at 3:49 PM, Dale Blount <dale(a)archlinux.org> wrote: >> On Thu, 2008-08-07 at 15:42 -0500, Aaron Griffin wrote: >>> On Thu, Aug 7, 2008 at 3:38 PM, David Rosenstrauch <darose(a)darose.net> wrote: >>> > I noticed that the pacman log doesn't get rotated. Seems like it should, as >>> > it could get rather big over time. Any particular reason why the package >>> > doesn't provide a /etc/logrotate.d/ script by default? >>> >>> I don't see a problem with that. Probably a good idea too. If you want >>> to provide a file for pacman, I'm sure Dan would be happy to include >>> it in the arch package. >> >> My only request here is that it saves at least a years worth of updates >> by default. If something breaks on my system and I don't notice it for >> a few months, I can't tell if an update broke it or not. >> >> I have a 828Kb pacman.log from a 5 year old install. Granted I don't >> -Syu as often as I should, but it still seems manageable at many times >> that on modern hardware. > > I'm actually with Dale here. I find it nice to go all the way back to > the "beginning of time" with my install so I can see exactly what may > have pulled in a now unneeded dep, etc. I just used this on my Eee > yesterday to remove unnecessary packages originally pulled in by > OpenOffice (hsqldb). I would rather old logs never get deleted; but > even more I would rather the file never get touched. > > There is a separate concern I have wanted to address for a while, and > that is the mixing of what was previously a pristine pacman.log with > the scriptlet messages. It is a great idea, but in practice, it makes > this file not near as concise as it once was. In an ideal world: > 1) pacman.log would return to only being upgrade/install/remove messages. > 2) another log file would be added that contained the verbose stuff. > pacman_messages.log or something. > 3) pacman.log never rotates/deletes. > 4) pacman_messages.log rotates/deletes. > > What do people think of this? While we're proposing ideas.... what about this: * new scriptlet function "message()", that just outputs text. * add a -Q operation to call the message() function. Then there's really no need for scriptlet logging at all.

4 4

[pacman-dev] -h/ --help option
by Matthias Gorissen 07 Aug '08

07 Aug '08

A user in the .de-forum pointed out that -h/ --help have the same output, even though the second version should display more info. Any thoughts on that?

5 5

[pacman-dev] [translation] Problems with DE-version
by Matthias Gorissen 07 Aug '08

07 Aug '08

So version 3.2.0 went to [core], and upon the next -Syu, pacman greeted me with the following message: :: pacman hat eine neuere Version von sich selbst gefunden. :: Möchtest du die momentan Handlung abbrechen :: und die neue pacman Version installieren? [Y/n]y WTF??? This is anything but German! Style and taste aside, there are at least three serious errors: 1. Pacman uses a colloquial form to address its user - this is logically inconsistent. All the rest of the De-translations use a more formal one (just to avoid ambivalent meanings - it is not about formality/politeness). 2. "momentan" should be "momentane" - it sounds like an "ery annoying error" in English. 3. [Y/n] should be [J/n]. If you type "y", pacman gets updated with all the other packages - not in before. Of course, I don't want to blame anyone in particular here. There is no-one to blame but me - I should have held these files under much closer surveillance. I was quite sure, however, that I had purged these lines, and that the changes were merged. So this an "ery annyoing error" indeed. For the sake of interest, this error was originally introduced here: http://projects.archlinux.org/?p=pacman.git;a=commitdiff;h=2b0c89b06a5cf68f… I didn't find any indication that this went over the ML (where I would have seen/checked it). Maybe a private mail to Dan? - I dunno, but it should not have been accepted this way. I thought I had cleaned up the entire issue here (but maybe I was wrong): http://projects.archlinux.org/?p=pacman.git;a=commitdiff;h=4ebad47ae7dff0f0… This one was merged, so I don't see why the the older version keeps popping up. I have provided another patch here: http://archlinux.org/pipermail/pacman-dev/2008-July/012427.html This one may not have been merged, but hey - there are at least 20 people subscribed to this who can checkout git-repos retrospectively 10 times better than I can, aren't they? So I'll write a patch for 3.2.1 (maybe this evening, if I stay sober enough). But despite my own short-comings, I wonder about organization. Maybe we could figure out a way to avoid such mistakes in the future. Best regards, Matt

2 3

[pacman-dev] added --config option to makepkg
by hoar heor 07 Aug '08

07 Aug '08

I added a --config=value option so you can use makepkg --config=safe and it will source the makepkg.safe file instead of makepkg.conf >From 4669853aac72506201af4d6e9ec79f080177d42f Mon Sep 17 00:00:00 2001 From: Imanol Celaya <ilcra1989(a)gmail.com> Date: Thu, 7 Aug 2008 13:15:27 +0200 Subject: [PATCH] adds --config=value option to use makepkg.value instead of makepkg.conf, as makepkg.safe or makepkg.blender Signed-off-by: Imanol Celaya <ilcra1989(a)gmail.com> --- scripts/makepkg.sh.in | 32 +++++++++++++++++++++++++------- 1 files changed, 25 insertions(+), 7 deletions(-) diff --git a/scripts/makepkg.sh.in b/scripts/makepkg.sh.in index 3604d10..267eb81 100644 --- a/scripts/makepkg.sh.in +++ b/scripts/makepkg.sh.in @@ -1084,6 +1084,7 @@ usage() { echo "$(gettext " -m, --nocolor Disable colorized output messages")" echo "$(gettext " -o, --nobuild Download and extract files only")" printf "$(gettext " -p <buildscript> Use an alternate build script (instead of '%s')")\n" "$BUILDSCRIPT" + echo "$(gettext " --config=value use /etc/makepkg.value file instead of makepkg.conf")" echo "$(gettext " -r, --rmdeps Remove installed dependencies after a successful build")" # fix flyspray feature request #2978 echo "$(gettext " -R, --repackage Repackage contents of pkg/ without building")" @@ -1126,14 +1127,31 @@ _PKGDEST=${PKGDEST} _SRCDEST=${SRCDEST} # Source makepkg.conf; fail if it is not found -if [ -r "$confdir/makepkg.conf" ]; then - source "$confdir/makepkg.conf" -else - error "$(gettext "%s not found.")" "$confdir/makepkg.conf" - plain "$(gettext "Aborting...")" - exit 1 # $E_CONFIG_ERROR +# if --config=value option found, use makepkg.value +CUSTOMCONFIG=0 +for ARG in $@; do +if [ ${ARG:0:9} = "--config=" ]; then + if [ -r "$confdir/makepkg.${ARG:9}" ]; then + plain "$(gettext "Using custom makepkg file: %s")" "$confdir/makepkg.${ARG:9}" + source "$confdir/makepkg.${ARG:9}" + CUSTOMCONFIG=1 + break + else + error "$(gettext "%s not found.")" "$confdir/makepkg.${ARG:9}" + plain "$(gettext "Aborting...")" + exit 1 # $E_CONFIG_ERROR +fi +if [ $CUSTOMCONFIG = 0 ]; then + if [ -r "$confdir/makepkg.conf" ]; then + source "$confdir/makepkg.conf" + else + error "$(gettext "%s not found.")" "$confdir/makepkg.conf" + plain "$(gettext "Aborting...")" + exit 1 # $E_CONFIG_ERROR + fi fi + # Source user-specific makepkg.conf overrides if [ -r ~/.makepkg.conf ]; then source ~/.makepkg.conf @@ -1166,7 +1184,7 @@ while true; do # Pacman Options --noconfirm) PACMAN_OPTS="$PACMAN_OPTS --noconfirm" ;; --noprogressbar) PACMAN_OPTS="$PACMAN_OPTS --noprogressbar" ;; - + # Makepkg Options --allsource) SOURCEONLY=2 ;; --asroot) ASROOT=1 ;; -- 1.5.6.5

3 7

[pacman-dev] [PATCH] dload.c : drop the specific handling of file: url.
by Xavier Chantry 06 Aug '08

06 Aug '08

Before commit fc48dc31, file:/// urls forced the use of the internal downloader (libdownload), because the default XferCommand, wget, does not handle them. We tried to move away from forcing usage of libdownload, so this commit implemented the handling of file:/// urls manually. However, this implementation is way too basic. It does not handle the progress bar, thus nothing at all appears in pacman's output when a file: repo is synchronized, or when a file is downloaded from a sync repo. Also, it is not able to detect when the repo is already up-to-date. When libdownload was used, both were handled. So it seems better to just drop this implementation for now. All users who use libdownload will get the much better file:// handling back. For the users of XferCommand, it will be more problematic, but they have several options: 1) switch to a downloader handling file:// (wget doesn't, but curl does for example) 2) drop the file:// repo, and set up light http or ftp servers instead. Consider that going that way would make this repo available for the whole local network, which can be useful 3) switch back to libdownload, which works perfectly for many users Signed-off-by: Xavier Chantry <shiningxc(a)gmail.com> --- lib/libalpm/dload.c | 15 --------------- 1 files changed, 0 insertions(+), 15 deletions(-) diff --git a/lib/libalpm/dload.c b/lib/libalpm/dload.c index b5f0b87..ef12646 100644 --- a/lib/libalpm/dload.c +++ b/lib/libalpm/dload.c @@ -340,21 +340,6 @@ cleanup: static int download(const char *url, const char *localpath, time_t mtimeold, time_t *mtimenew) { int ret; - const char *proto = "file://"; - int len = strlen(proto); - if(strncmp(url, proto, len) == 0) { - /* we can simply grab an absolute path from the file:// url by starting - * our path at the char following the proto (the root '/') - */ - const char *sourcefile = url + len; - const char *filename = get_filename(url); - char *destfile = get_destfile(localpath, filename); - - ret = _alpm_copyfile(sourcefile, destfile); - FREE(destfile); - /* copyfile returns 1 on failure, we want to return -1 on failure */ - return(ret ? -1 : 0); - } /* We have a few things to take into account here. * 1. If we have both internal/external available, choose based on -- 1.5.6.4

2 1

[pacman-dev] [GIT] The official pacman repository branch, master, updated. v3.2.0-4-g72c5a29
by dan＠archlinux.org 06 Aug '08

06 Aug '08

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 72c5a298a3ee0f18019010ef2eb43da654ec25f8 (commit) via 4476598e4e128f4595d5383ecb51a9576a447b5b (commit) via 9bc799ec7b1718e6d90ceedd5e08672068e26e10 (commit) via 692ea72822e6cf11930fc8f85afd23c1b65f1e8e (commit) from fabf7ad6c4d9e0a45bf0957d93d810a361db7418 (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 72c5a298a3ee0f18019010ef2eb43da654ec25f8 Author: Xavier Chantry <shiningxc(a)gmail.com> Date: Mon Aug 4 01:16:08 2008 +0200 Avoid double slashes in URLs given to libdownload. If a Server specified in pacman.conf had a trailing slash, libalpm ended up building URLs with double slashes, and this broke libdownload with errors like the following one : error: failed retrieving file 'redland-1.0.8-1-i686.pkg.tar.gz' from 192.168.0.90 : Command okay So the public function alpm_db_set_server will make sure to remove the trailing slash of servers. For the private function _alpm_download_single_file, I only added a comment. Signed-off-by: Xavier Chantry <shiningxc(a)gmail.com> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 4476598e4e128f4595d5383ecb51a9576a447b5b Author: Xavier Chantry <shiningxc(a)gmail.com> Date: Sun Aug 3 20:03:05 2008 +0200 dload.c : drop the specific handling of file: url. Before commit fc48dc31, file:/// urls forced the use of the internal downloader (libdownload), because the default XferCommand, wget, does not handle them. We tried to move away from forcing usage of libdownload, so this commit implemented the handling of file:/// urls manually. However, this implementation is way too basic. It does not handle the progress bar, thus nothing at all appears in pacman's output when a file: repo is synchronized, or when a file is downloaded from a sync repo. Also, it is not able to detect when the repo is already up-to-date. When libdownload was used, both were handled. It seems better to just drop this implementation for now. All users who use libdownload will get the much better file:// handling back. For the users of XferCommand, it will be more problematic, but they have several options: 1) Switch to a downloader handling file:// (wget doesn't, but curl does for example). 2) Drop the file:// repo, and set up light http or ftp servers instead. Consider that going that way would make this repo available for the whole local network, which can be useful. 3) Switch back to libdownload, which works perfectly for many users. Signed-off-by: Xavier Chantry <shiningxc(a)gmail.com> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 9bc799ec7b1718e6d90ceedd5e08672068e26e10 Author: Xavier Chantry <shiningxc(a)gmail.com> Date: Fri Aug 1 08:29:46 2008 +0200 makepkg: fix download functions with weird urls. This fixes FS#11076. 1) quote the url in get_downloadclient 2) only enable nullglob where it is needed You can see in 7fc306cd41 that nullglob was only enabled for one part, and that it already caused other problems, which were fixed in 7ff5a917fd0. Thanks to Henning Garus for pointing out that nullglob was problematic with urls containing expansion char like '?'. 3) change get_downloadcmd which displayed the download command line to download_file which actually executes the download. It seems nicer that way. Signed-off-by: Xavier Chantry <shiningxc(a)gmail.com> Signed-off-by: Dan McGee <dan(a)archlinux.org> commit 692ea72822e6cf11930fc8f85afd23c1b65f1e8e Author: Dan McGee <dan(a)archlinux.org> Date: Wed Jul 30 21:53:02 2008 -0500 makepkg: Clarify usage instructions for --allsource/--source Signed-off-by: Dan McGee <dan(a)archlinux.org> ----------------------------------------------------------------------- Summary of changes: doc/makepkg.8.txt | 14 +++++++++++--- lib/libalpm/db.c | 16 +++++++++++++--- lib/libalpm/dload.c | 16 +--------------- scripts/makepkg.sh.in | 29 ++++++++++++++++------------- 4 files changed, 41 insertions(+), 34 deletions(-) hooks/post-receive -- The official pacman repository

2 1

[pacman-dev] [PATCH] Attempt to idiot-proof making and refreshing docs
by Dan McGee 06 Aug '08

06 Aug '08

I mess this up more often than not, and maybe this will do the trick. Remove the --enable-asciidoc option as it has been superseded by the --disable-doc option in usefulness. If you want to skip building docs, you skip building all docs which is much easier when it comes to ensuring the make 'dist' and 'distcheck' targets will always build the manpages and always build the most up to date manpages. Developers shouldn't be affected in their normal builds, nor should end users of the source tarball. Signed-off-by: Dan McGee <dan(a)archlinux.org> --- Makefile.am | 3 +++ configure.ac | 23 ----------------------- doc/Makefile.am | 22 ++++++++++++---------- 3 files changed, 15 insertions(+), 33 deletions(-) diff --git a/Makefile.am b/Makefile.am index ea3b0e0..ed9a34d 100644 --- a/Makefile.am +++ b/Makefile.am @@ -3,6 +3,9 @@ if WANT_DOC SUBDIRS += doc endif +# Make sure we test and build manpages when doing distcheck +DISTCHECK_CONFIGURE_FLAGS = --enable-doc --disable-git-version + # Some files automatically included, so they aren't specified below: # AUTHORS, COPYING, NEWS, README EXTRA_DIST = HACKING diff --git a/configure.ac b/configure.ac index 9f4209b..9c5ed23 100644 --- a/configure.ac +++ b/configure.ac @@ -103,11 +103,6 @@ AC_ARG_ENABLE(doxygen, AS_HELP_STRING([--enable-doxygen], [build your own API docs via Doxygen]), [wantdoxygen=$enableval], [wantdoxygen=no]) -# Help line for asciidoc -AC_ARG_ENABLE(asciidoc, - AS_HELP_STRING([--enable-asciidoc], [build your own manpages with Asciidoc]), - [wantasciidoc=$enableval], [wantasciidoc=no]) - # Help line for debug AC_ARG_ENABLE(debug, AS_HELP_STRING([--enable-debug], [enable debugging support]), @@ -267,23 +262,6 @@ else fi AM_CONDITIONAL(USE_DOXYGEN, test "x$usedoxygen" = "xyes") -# Check for asciidoc support and status -AC_CHECK_PROGS([ASCIIDOC], [asciidoc]) -AC_MSG_CHECKING([for asciidoc]) -if test "x$wantasciidoc" = "xyes" ; then - if test $ASCIIDOC ; then - AC_MSG_RESULT([yes]) - useasciidoc=yes - else - AC_MSG_RESULT([no, asciidoc missing]) - useasciidoc=no - fi -else - AC_MSG_RESULT([no, disabled by configure]) - useasciidoc=no -fi -AM_CONDITIONAL(USE_ASCIIDOC, test "x$useasciidoc" = "xyes") - # Enable or disable debug code AC_MSG_CHECKING(for debug mode request) if test "x$debug" = "xyes" ; then @@ -381,7 +359,6 @@ ${PACKAGE_NAME}: Run make in doc/ dir : ${wantdoc} Use download library : ${internaldownload} Doxygen support : ${usedoxygen} - Asciidoc support : ${useasciidoc} debug support : ${debug} " diff --git a/doc/Makefile.am b/doc/Makefile.am index 476a21c..cce0a71 100644 --- a/doc/Makefile.am +++ b/doc/Makefile.am @@ -15,6 +15,7 @@ ASCIIDOC_MANS = \ DOXYGEN_MANS = $(wildcard man3/*.3) EXTRA_DIST = \ + asciidoc.conf \ pacman.8.txt \ makepkg.8.txt \ repo-add.8.txt \ @@ -29,8 +30,12 @@ EXTRA_DIST = \ $(DOXYGEN_MANS) # Files that should be removed, but which Automake does not know. -MOSTLYCLEANFILES = *.xml -MAINTAINERCLEANFILES = $(ASCIIDOC_MANS) +MOSTLYCLEANFILES = *.xml $(ASCIIDOC_MANS) + +# Ensure manpages are fresh when building a dist tarball +dist-hook: + $(MAKE) $(AM_MAKEFLAGS) clean + $(MAKE) $(AM_MAKEFLAGS) all if USE_GIT_VERSION GIT_VERSION := $(shell sh -c 'git describe --abbrev=4 | sed s/^v//')-dirty @@ -39,20 +44,18 @@ else REAL_PACKAGE_VERSION = $(PACKAGE_VERSION) endif - man_MANS = dist_man_MANS = $(ASCIIDOC_MANS) repo-remove.8 if USE_DOXYGEN man_MANS += $(DOXYGEN_MANS) -all: doxygen.in +all-local: doxygen.in doxygen.in: $(DOXYGEN) $(srcdir)/Doxyfile endif -if USE_ASCIIDOC ASCIIDOC_OPTS = \ -f asciidoc.conf \ -a pacman_version="$(REAL_PACKAGE_VERSION)" \ @@ -64,11 +67,10 @@ A2X_OPTS = \ --xsltproc-opts='-param man.endnotes.list.enabled 0' \ --xsltproc-opts='-param man.endnotes.are.numbered 0' -$(ASCIIDOC_MANS): - a2x $(A2X_OPTS) --asciidoc-opts="$(ASCIIDOC_OPTS)" $@.txt - # These rules are due to the includes and files of the asciidoc text $(ASCIIDOC_MANS): asciidoc.conf footer.txt + a2x $(A2X_OPTS) --asciidoc-opts="$(ASCIIDOC_OPTS)" $@.txt + pacman.8: pacman.8.txt makepkg.8: makepkg.8.txt repo-add.8: repo-add.8.txt @@ -78,7 +80,7 @@ pacman.conf.5: pacman.conf.5.txt libalpm.3: libalpm.3.txt # this one is just a symlink repo-remove.8: repo-add.8 - ln -s repo-add.8 repo-remove.8 -endif + rm -f repo-remove.8 + $(LN_S) repo-add.8 repo-remove.8 # vim:set ts=2 sw=2 noet: -- 1.5.6.4

1 0

[pacman-dev] [PATCH] pactest: remove need for pause when running tests
by Dan McGee 04 Aug '08

04 Aug '08

Instead of pausing 1.5 seconds on tests that check file mtimes, change the mtimes to something in the far past so we can immediately tell if a file was modified and/or touched. This saves a decent amount of time on the upgrade tests which often check mtimes. 355 was a completely arbitrary time value, don't ask me why I picked it. Signed-off-by: Dan McGee <dan(a)archlinux.org> --- pactest/pmenv.py | 8 +++++--- pactest/pmfile.py | 8 ++++++++ 2 files changed, 13 insertions(+), 3 deletions(-) diff --git a/pactest/pmenv.py b/pactest/pmenv.py index e9d0b4c..cc433dd 100755 --- a/pactest/pmenv.py +++ b/pactest/pmenv.py @@ -71,9 +71,11 @@ class pmenv: t.generate() # Hack for mtimes consistency for i in t.rules: - if i.rule.find("MODIFIED") != -1: - time.sleep(1.5) - break + if i.rule.find("FILE_MODIFIED") != -1: + [test, arg] = i.rule.split("=") + for f in t.files: + if f.name == arg: + f.resettimes() t.run(self.pacman) diff --git a/pactest/pmfile.py b/pactest/pmfile.py index 21baf01..ab4aa2c 100755 --- a/pactest/pmfile.py +++ b/pactest/pmfile.py @@ -56,6 +56,14 @@ class pmfile: return retval + def resettimes(self): + """ + """ + + filename = os.path.join(self.root, self.name) + os.utime(filename, (355, 355)) + self.mtime = getmtime(filename) + vprint("\tmtime reset (%s)" % self.name) if __name__ == "__main__": f = pmfile("/tmp", "foobar") -- 1.5.6.3

3 5

[pacman-dev] [PATCH] Change download functions to handle funny urls
by Henning Garus 01 Aug '08

01 Aug '08

Stop urls containing & or ? from breaking get_downloadclient() and get_downloadcmd(). Fixes FS#11076 Signed-off-by: Henning Garus <henning.garus(a)gmail.com> --- scripts/makepkg.sh.in | 8 ++++++-- 1 files changed, 6 insertions(+), 2 deletions(-) diff --git a/scripts/makepkg.sh.in b/scripts/makepkg.sh.in index 041b460..38364a6 100644 --- a/scripts/makepkg.sh.in +++ b/scripts/makepkg.sh.in @@ -256,7 +256,7 @@ in_array() { get_downloadclient() { # $1 = url with valid protocol prefix local url=$1 - local proto=$(echo $netfile | sed 's|://.*||') + local proto=$(echo "$url" | sed 's|://.*||') # loop through DOWNLOAD_AGENTS variable looking for protocol local i @@ -293,7 +293,8 @@ get_downloadcmd() { local file=$3 if echo "$dlagent" | grep -q "%u" ; then - local dlcmd=$(echo "$dlagent" | sed "s|%o|$file.part|" | sed "s|%u|$netfile|") + #use awk because of possible & in $netfile (FS#11076) + local dlcmd=$(echo "$dlagent" | sed "s|%o|$file.part|" | awk -F "%u" -v nf="$netfile" '{ print $1 nf $2}') else local dlcmd="$dlagent $netfile" fi @@ -443,7 +444,10 @@ download_sources() { msg2 "$(gettext "Downloading %s...")" "$file" # fix flyspray bug #3289 local ret=0 + #disable expansion of possible ? in $netfile (FS#11076) + set -f $(get_downloadcmd "$dlclient" "$netfile" "$file") || ret=$? + set +f if [ $ret -gt 0 ]; then error "$(gettext "Failure while downloading %s")" "$file" plain "$(gettext "Aborting...")" -- 1.5.6.4

2 1