implement POC client-side merging of pdiffs via apt-file

The idea of pdiffs is to avoid downloading the hole file by patching the
existing index. This works very well, but becomes slow if a lot of
patches needs to be applied to reconstruct an up-to-date index and in
recent years more and more dinstall (or similar) runs are executed
creating more and more pdiffs in the same amount of time, so pdiffs
became less useful.

The solution is simple: Reduce the amount of patches (which are very
small) which need to be applied on top of the index we have available
(which is usually pretty big).

This can be done in two ways: Either merge the patches on the
server-side so that the client has to download only one patch or the
patches are all downloaded and merged on the client-side.

The first needs a client who is doing one step at a time who can also
skip patches if it needs (APT supports this for a long time now).

The later is implemented by this commit, but depends on the server NOT
merging the patches and the patches being in a strict order in which no
patch is skipped.

This is traditionally the case for dak, but other repository creators
support merging – e.g. reprepro (which helpfully adds a flag indicating
that the patches are merged). To support both or even mixes a client
needs more information which isn't available for now.

This POC uses the external diffindex-rred included in apt-file to
do the heavy lifting of merging & applying all patches in one pass,
hence to test this feature apt-file needs to be installed.

1 files changed, 127 insertions, 20 deletions

diff --git a/test/integration/test-pdiff-usage b/test/integration/test-pdiff-usage
index ac0563b7f..5a06e0ccb 100755
--- a/test/integration/test-pdiff-usage
+++ b/test/integration/test-pdiff-usage
@@ -5,39 +5,146 @@ TESTDIR=$(readlink -f $(dirname $0))
 . $TESTDIR/framework
 
 setupenvironment
-configarchitecture "i386"
+configarchitecture 'i386'
 
 buildaptarchive
 setupflataptarchive
 changetowebserver
-signreleasefiles
-testsuccess aptget update
 
-testnopackage newstuff
 PKGFILE="${TESTDIR}/$(echo "$(basename $0)" | sed 's#^test-#Packages-#')"
-testequal "$(cat ${PKGFILE})
+
+echo '#!/bin/sh
+touch merge-was-used
+/usr/bin/diffindex-rred "$@"' > extrred
+chmod +x extrred
+echo 'Dir::Bin::rred "./extrred";' > rootdir/etc/apt/apt.conf.d/99rred
+
+wasmergeused() {
+	testsuccess aptget update "$@"
+	msgtest 'Check if the right pdiff merger was used'
+	if [ -e ./merge-was-used ]; then
+		rm -f ./merge-was-used
+		if echo "$*" | grep -q -- '-o Acquire::PDiffs::Merge=1'; then
+			msgpass
+		else
+			msgfail "Merge shouldn't have been used, but was"
+		fi
+	elif echo "$*" | grep -q -- '-o Acquire::PDiffs::Merge=1'; then
+		msgfail "Merge should have been used, but wasn't"
+	else
+		msgpass
+	fi
+}
+
+testrun() {
+	# setup the base
+	find aptarchive -name 'Packages*' -type f -delete
+	cp ${PKGFILE} aptarchive/Packages
+	compressfile 'aptarchive/Packages'
+	generatereleasefiles
+	signreleasefiles
+	rm -rf aptarchive/Packages.diff rootdir/var/lib/apt/lists
+	testsuccess aptget update "$@"
+	cp -a rootdir/var/lib/apt/lists rootdir/var/lib/apt/lists-bak
+	testnopackage newstuff
+	testequal "$(cat ${PKGFILE})
 " aptcache show apt oldstuff
 
-cp ${PKGFILE}-new aptarchive/Packages
-compressfile 'aptarchive/Packages'
-rm -rf aptarchive/Packages.diff
-mkdir -p aptarchive/Packages.diff
-PATCHFILE="aptarchive/Packages.diff/$(date +%Y-%m-%d-%H%M.%S)"
-diff -e ${PKGFILE} ${PKGFILE}-new > ${PATCHFILE} || true
-cat $PATCHFILE | gzip > ${PATCHFILE}.gz
-PATCHINDEX="aptarchive/Packages.diff/Index"
-echo "SHA1-Current: $(sha1sum ${PKGFILE}-new | cut -d' ' -f 1) $(stat -c%s ${PKGFILE}-new)
+	# apply with one patch
+	cp ${PKGFILE}-new aptarchive/Packages
+	compressfile 'aptarchive/Packages'
+	mkdir -p aptarchive/Packages.diff
+	PATCHFILE="aptarchive/Packages.diff/$(date +%Y-%m-%d-%H%M.%S)"
+	diff -e ${PKGFILE} ${PKGFILE}-new > ${PATCHFILE} || true
+	cat $PATCHFILE | gzip > ${PATCHFILE}.gz
+	PATCHINDEX='aptarchive/Packages.diff/Index'
+	echo "SHA1-Current: $(sha1sum ${PKGFILE}-new | cut -d' ' -f 1) $(stat -c%s ${PKGFILE}-new)
 SHA1-History:
  9f4148e06d7faa37062994ff10d0c842d7017513 33053002 2010-08-18-2013.28
  $(sha1sum $PKGFILE | cut -d' ' -f 1) $(stat -c%s $PKGFILE) $(basename $PATCHFILE)
 SHA1-Patches:
  7651fc0ac57cd83d41c63195a9342e2db5650257 19722 2010-08-18-0814.28
  $(sha1sum $PATCHFILE | cut -d' ' -f 1) $(stat -c%s $PATCHFILE) $(basename $PATCHFILE)" > $PATCHINDEX
-generatereleasefiles '+1hour'
-signreleasefiles
-find aptarchive -name 'Packages*' -type f -delete
-testsuccess aptget update
+	generatereleasefiles '+1hour'
+	signreleasefiles
+	find aptarchive -name 'Packages*' -type f -delete
+	wasmergeused "$@"
+	testnopackage oldstuff
+	testequal "$(cat ${PKGFILE}-new)
+" aptcache show apt newstuff
 
-testnopackage oldstuff
-testequal "$(cat ${PKGFILE}-new)
+	# index is already up-to-date
+	find rootdir/var/lib/apt/lists -name '*.IndexDiff' -type f -delete
+	testsuccess aptget update "$@"
+	testequal "$(cat ${PKGFILE}-new)
 " aptcache show apt newstuff
+
+	# apply with two patches
+	cp ${PKGFILE}-new aptarchive/Packages
+	echo '
+Package: futurestuff
+Version: 1.0
+Architecture: i386
+Maintainer: Joe Sixpack <joe@example.org>
+Installed-Size: 202
+Filename: pool/futurestuff_1.0_i386.deb
+Size: 202200
+MD5sum: 311aeeaaae5ba33aff1ceaf3e1f76671
+SHA1: 3c695e028f7a1ae324deeaae5ba332desa81088c
+SHA256: b46fd154615edaae5ba33c56a5cc0e7deaef23e2da3e4f129727fd660f28f050
+Description: some cool and shiny future stuff
+ This package will appear in the next next mirror update
+Description-md5: d5f89fbbc2ce34c455dfee9b67d82b6b' >> aptarchive/Packages
+
+	compressfile 'aptarchive/Packages'
+	PATCHFILE2="aptarchive/Packages.diff/$(date -d 'now + 1hour' '+%Y-%m-%d-%H%M.%S')"
+	diff -e ${PKGFILE}-new aptarchive/Packages > ${PATCHFILE2} || true
+	cat $PATCHFILE2 | gzip > ${PATCHFILE2}.gz
+	echo "SHA1-Current: $(sha1sum aptarchive/Packages | cut -d' ' -f 1) $(stat -c%s aptarchive/Packages)
+SHA1-History:
+ 9f4148e06d7faa37062994ff10d0c842d7017513 33053002 2010-08-18-2013.28
+ $(sha1sum ${PKGFILE} | cut -d' ' -f 1) $(stat -c%s ${PKGFILE}) $(basename ${PATCHFILE})
+ $(sha1sum ${PKGFILE}-new | cut -d' ' -f 1) $(stat -c%s ${PKGFILE}-new) $(basename ${PATCHFILE2})
+SHA1-Patches:
+ 7651fc0ac57cd83d41c63195a9342e2db5650257 19722 2010-08-18-0814.28
+ $(sha1sum $PATCHFILE | cut -d' ' -f 1) $(stat -c%s $PATCHFILE) $(basename $PATCHFILE)
+ $(sha1sum ${PATCHFILE2} | cut -d' ' -f 1) $(stat -c%s ${PATCHFILE2}) $(basename ${PATCHFILE2})" > $PATCHINDEX
+	generatereleasefiles '+2hour'
+	signreleasefiles
+	cp -a aptarchive/Packages Packages-future
+	find aptarchive -name 'Packages*' -type f -delete
+	rm -rf rootdir/var/lib/apt/lists
+	cp -a rootdir/var/lib/apt/lists-bak rootdir/var/lib/apt/lists
+	wasmergeused "$@"
+	testnopackage oldstuff
+	testequal "$(cat Packages-future)
+" aptcache show apt newstuff futurestuff
+
+	# patch applying fails, but successful fallback
+	rm -rf rootdir/var/lib/apt/lists
+	cp -a rootdir/var/lib/apt/lists-bak rootdir/var/lib/apt/lists
+	cp ${PKGFILE}-new aptarchive/Packages
+	compressfile 'aptarchive/Packages'
+	mkdir -p aptarchive/Packages.diff
+	PATCHFILE="aptarchive/Packages.diff/$(date +%Y-%m-%d-%H%M.%S)"
+	diff -e ${PKGFILE} ${PKGFILE}-new > ${PATCHFILE} || true
+	PATCHINDEX='aptarchive/Packages.diff/Index'
+	echo "SHA1-Current: $(sha1sum ${PKGFILE}-new | cut -d' ' -f 1) $(stat -c%s ${PKGFILE}-new)
+SHA1-History:
+ 9f4148e06d7faa37062994ff10d0c842d7017513 33053002 2010-08-18-2013.28
+ $(sha1sum $PKGFILE | cut -d' ' -f 1) $(stat -c%s $PKGFILE) $(basename $PATCHFILE)
+SHA1-Patches:
+ 7651fc0ac57cd83d41c63195a9342e2db5650257 19722 2010-08-18-0814.28
+ $(sha1sum $PATCHFILE | cut -d' ' -f 1) $(stat -c%s $PATCHFILE) $(basename $PATCHFILE)" > $PATCHINDEX
+	echo 'I am Mallory and I change files' >> $PATCHFILE
+	cat $PATCHFILE | gzip > ${PATCHFILE}.gz
+	generatereleasefiles '+1hour'
+	signreleasefiles
+	testsuccess aptget update "$@"
+	testnopackage oldstuff
+	testequal "$(cat ${PKGFILE}-new)
+" aptcache show apt newstuff
+}
+
+testrun -o Debug::pkgAcquire::Diffs=1 -o Debug::pkgAcquire::rred=1 -o Acquire::PDiffs::Merge=0
+testrun -o Debug::pkgAcquire::Diffs=1 -o Debug::pkgAcquire::rred=1 -o Acquire::PDiffs::Merge=1