summaryrefslogtreecommitdiff
path: root/test
AgeCommit message (Collapse)Author
2012-06-14add extra paranoia against subkey attacks (and a regression test), LP: ↵Michael Vogt
#1013128, thanks to jdstrand and mdeslaur and Geori Guinski
2012-06-14update to reflect the latest ubuntu keyMichael Vogt
2012-06-11Merge version 0.9.6 from DebianSteve Langasek
2012-06-11add a few tests to check that arch-specific qualifiers workDavid Kalnischkies
2012-06-11check build-dep candidate if install is forbiddenDavid Kalnischkies
2012-06-10* apt-pkg/aptconfiguration.cc:David Kalnischkies
- if APT::Languages=none save "none" in allCodes so that the detected configuration is cached as intended (Closes: #674690, LP: #1004947)
2012-06-10* cmdline/apt-get.cc:Daniel Hartwig
- print URIs for all changelogs in case of --print-uris, thanks to Daniel Hartwig for the patch! (Closes: #674897)
2012-06-09fix segfault with empty LongOpt in --no-* branchDavid Kalnischkies
2012-06-09* apt-pkg/contrib/cmdline.cc:Daniel Hartwig
- apply patch from Daniel Hartwig to fix a segfault in case the LongOpt is empty (Closes: #676331)
2012-05-22merged (and resolved a bunch of conflicts) of the debian-sid branchMichael Vogt
2012-05-20* apt-pkg/aptconfiguration.cc:David Kalnischkies
- longcode Translation files are saved with encoded underscore, so make sure to pick these files up as well for Acquire::Languages
2012-05-19* apt-pkg/packagemanager.cc:David Kalnischkies
- do not run into loop on new-pre-depends-breaks (Closes: #673536)
2012-05-16* cmdline/apt-get.cc:David Kalnischkies
- use the host architecture, not the build architecture for matching of [architecture restrictions] in Build-Depends (Closes: #672927) *
2012-05-14adapt manual testcase to check redirects of dists/, tooDavid Kalnischkies
2012-05-14 - separate manpages from the rest of the doc buildingDavid Kalnischkies
- make apt and apt-utils packages depend on manpages instead of full doc
2012-05-10add a small wrapper to call gdb in the framework settingDavid Kalnischkies
2012-05-02add a simple test for HashSumValue classesDavid Kalnischkies
2012-04-23/dev/null is a special absolute path as it has no subdirectoriesDavid Kalnischkies
2012-04-23normalize a bit by replacing // and /./ with / in FindFileDavid Kalnischkies
2012-04-23* apt-pkg/depcache.cc:David Kalnischkies
- clearly separate 'positive' and 'negative' dependencies and their upgrade-resolution tries in MarkInstall and especially don't treat Conflicts differently compared to Breaks here
2012-04-20merged r2204 from lp:~donkult/apt/sidMichael Vogt
2012-04-20* apt-pkg/cacheset.cc:David Kalnischkies
- actually return to the fallback modifier if we have detected we should for packagenames which look like modifiers (Closes: #669591)
2012-04-19add a very simple testcase for the lp#985852 fix in the previous commitDavid Kalnischkies
2012-04-19ignore the order used for the 'apt' package as long as it is valid either wayDavid Kalnischkies
2012-04-17ensure that not --assert-multi-arch supporting dpkg's do not generateDavid Kalnischkies
output in the testcases by redirecting to /dev/null
2012-04-12merged r1972 from lp:~paolorotolo/apt/fix-for-967393Michael Vogt
2012-04-11fix the remaining lzma calls with xz --format=lzma in the testcasesDavid Kalnischkies
2012-04-11* cmdline/apt-get.cc:David Kalnischkies
- print list of autoremoves in alphabetical order (Closes: #639008)
2012-04-11use xz-utils in the testcases instead of lzma and ensure that we reallyDavid Kalnischkies
ignore the presents (or absence) of lzma if we decided to use xz
2012-04-05For cross-build-dependencies, Architecture: all packages should beSteve Langasek
treated as implicitly Multi-Arch: foreign, because either they *are* M-A: foreign when used as a build-dependency, or they need to be updated to not be Architecture: all; and since cross-build-deps are new functionality in apt, we can safely make this change without breaking existing systems. Closes: #666772.
2012-04-05* cmdline/apt-get.cc:Bogdan Purcareata
- distinguish information about 'apt-get autoremove' based on the number of auto-removed packages both before and after the list of packages (Closes: #665833)
2012-03-22use the variables as the rest of the tests to avoid escapes in the outputDavid Kalnischkies
2012-03-22* apt-inst/database.{cc,h}, apt-inst/deb/dpkgdb.{cc,h}:David Kalnischkies
- drop instead of fix as it is only needed if you want to reimplement dpkg and comes straight from the beginning of last decade (Closes: #663372) * apt-inst/deb/debfile.cc: - {Extract,Merge}Control() is another instance of "lets reimplement dpkg" so shot of this code before someone ends up using this…
2012-03-21do not update po and pot files in the process of the build as thisDavid Kalnischkies
causes timestamp changes for the mo files which therefore can't be refcounted by dpkg for your M-A: same packages (Closes: #659333, LP: #924628) The commit also enables a top-level 'make update-po' and does all the needed changes to let this work now that update-po might be called in a freshly checkout tree
2012-03-20for these helpers we don't need too much checks,David Kalnischkies
but no warnings from gcc helps
2012-03-20update the versionnumber of the faked dpkg in the testcases for multiarchDavid Kalnischkies
and add a comment about the need of this number
2012-03-13cherry pick r2226Michael Vogt
2012-03-13cherry pick 2217.1.22Michael Vogt
2012-03-13now that stdout/stdlog are used consistently, the message order changedDavid Kalnischkies
2012-03-13recheck dependencies in SmartUnpack after a change, tooDavid Kalnischkies
2012-03-12ensure that the fullname of a package is displayed in the debug outputDavid Kalnischkies
2012-03-12improve the testcode so that we do not depend on the order of unpackingDavid Kalnischkies
of specific packages as long as the order is okay in general
2012-03-12now that the package name 'apt' is a magic word enabling essential statusDavid Kalnischkies
for said package i can't use it anymore in the testcase, so use another name
2012-03-06* merged from the debian-sid branch, most notably:Michael Vogt
- Correct fi translation for hash sum mismatches (LP: #420403) - remove 'old' InRelease file if we can't get a new one before proceeding with Release.gpg to avoid the false impression of a still trusted repository by a (still present) old InRelease file. Thanks to Simon Ruderich for reporting this issue! (CVE-2012-0214) * apt-pkg/acquire-item.cc: - remove 'old' InRelease file if we can't get a new one before proceeding with Release.gpg to avoid the false impression of a still trusted repository by a (still present) old InRelease file. Thanks to Simon Ruderich for reporting this issue! (CVE-2012-0214) * French: replace "étiquetage" by "épinglage" for "pinning"
2012-03-06* apt-pkg/acquire-item.cc:Michael Vogt
- remove 'old' InRelease file if we can't get a new one before proceeding with Release.gpg to avoid the false impression of a still trusted repository by a (still present) old InRelease file. Thanks to Simon Ruderich for reporting this issue! (CVE-2012-0214)
2012-03-03testcase for CVE-2012-0214Simon Ruderich
2012-03-03* apt-pkg/packagemanager.cc:David Kalnischkies
- do not try to a void a breaks if the broken package pre-depends on the breaker, but let dpkg auto-deconfigure it
2012-03-03allow msgtest to be used with only one parameterDavid Kalnischkies
2012-02-27fix the string in the testcasesDavid Kalnischkies
2012-02-18improve 'error' message for packages which are only referencedDavid Kalnischkies
e.g. in a Depends line and are now requested for removal