share description list between "same" versions (LP: #868977)

3 files changed, 80 insertions, 17 deletions

diff --git a/apt-pkg/pkgcachegen.cc b/apt-pkg/pkgcachegen.cc
index 5147d7547..3545517fe 100644
--- a/apt-pkg/pkgcachegen.cc
+++ b/apt-pkg/pkgcachegen.cc
@@ -38,6 +38,9 @@
 typedef vector<pkgIndexFile *>::iterator FileIterator;
 template <typename Iter> std::vector<Iter*> pkgCacheGenerator::Dynamic<Iter>::toReMap;
 
+bool IsDuplicateDescription(pkgCache::DescIterator Desc,
+			    MD5SumValue const &CurMd5, std::string const &CurLang);
+
 // CacheGenerator::pkgCacheGenerator - Constructor			/*{{{*/
 // ---------------------------------------------------------------------
 /* We set the dirty flag and make sure that is written to the disk */
@@ -242,8 +245,11 @@ bool pkgCacheGenerator::MergeList(ListParser &List,
 bool pkgCacheGenerator::MergeListGroup(ListParser &List, std::string const &GrpName)
 {
    pkgCache::GrpIterator Grp = Cache.FindGrp(GrpName);
+   // a group has no data on it's own, only packages have it but these
+   // stanzas like this come from Translation- files to add descriptions,
+   // but without a version we don't need a description for it…
    if (Grp.end() == true)
-      return _error->Error("Information merge for non-existent group %s requested", GrpName.c_str());
+      return true;
    Dynamic<pkgCache::GrpIterator> DynGrp(Grp);
 
    pkgCache::PkgIterator Pkg;
@@ -268,6 +274,7 @@ bool pkgCacheGenerator::MergeListPackage(ListParser &List, pkgCache::PkgIterator
 
    // Find the right version to write the description
    MD5SumValue CurMd5 = List.Description_md5();
+   std::string CurLang = List.DescriptionLanguage();
 
    for (Ver = Pkg.VersionList(); Ver.end() == false; ++Ver)
    {
@@ -275,15 +282,9 @@ bool pkgCacheGenerator::MergeListPackage(ListParser &List, pkgCache::PkgIterator
       Dynamic<pkgCache::DescIterator> DynDesc(Desc);
       map_ptrloc *LastDesc = &Ver->DescriptionList;
 
-
       // don't add a new description if we have one for the given
       // md5 && language
-      bool duplicate = false;
-      for ( ; Desc.end() == false; ++Desc)
-	 if (MD5SumValue(Desc.md5()) == CurMd5 &&
-	     Desc.LanguageCode() == List.DescriptionLanguage())
-	       duplicate=true;
-      if (duplicate == true)
+      if (IsDuplicateDescription(Desc, CurMd5, CurLang) == true)
 	 continue;
 
       for (Desc = Ver.DescriptionList();
@@ -295,7 +296,7 @@ bool pkgCacheGenerator::MergeListPackage(ListParser &List, pkgCache::PkgIterator
 
 	 // Add new description
 	 void const * const oldMap = Map.Data();
-	 map_ptrloc const descindex = NewDescription(Desc, List.DescriptionLanguage(), CurMd5, *LastDesc);
+	 map_ptrloc const descindex = NewDescription(Desc, CurLang, CurMd5, *LastDesc);
 	 if (oldMap != Map.Data())
 	    LastDesc += (map_ptrloc*) Map.Data() - (map_ptrloc*) oldMap;
 	 *LastDesc = descindex;
@@ -391,16 +392,34 @@ bool pkgCacheGenerator::MergeListVersion(ListParser &List, pkgCache::PkgIterator
       return true;
    }
 
-   /* Record the Description data. Description data always exist in
-      Packages and Translation-* files. */
+   /* Record the Description (it is not translated) */
+   MD5SumValue CurMd5 = List.Description_md5();
+   if (CurMd5.Value().empty() == true)
+      return true;
+   std::string CurLang = List.DescriptionLanguage();
+
+   /* Before we add a new description we first search in the group for
+      a version with a description of the same MD5 - if so we reuse this
+      description group instead of creating our own for this version */
+   pkgCache::GrpIterator Grp = Pkg.Group();
+   for (pkgCache::PkgIterator P = Grp.PackageList();
+	P.end() == false; P = Grp.NextPkg(P))
+   {
+      for (pkgCache::VerIterator V = P.VersionList();
+	   V.end() == false; ++V)
+      {
+	 if (IsDuplicateDescription(V.DescriptionList(), CurMd5, "") == false)
+	    continue;
+	 Ver->DescriptionList = V->DescriptionList;
+	 return true;
+      }
+   }
+
+   // We haven't found reusable descriptions, so add the first description
    pkgCache::DescIterator Desc = Ver.DescriptionList();
    Dynamic<pkgCache::DescIterator> DynDesc(Desc);
    map_ptrloc *LastDesc = &Ver->DescriptionList;
 
-   // Skip to the end of description set
-   for (; Desc.end() == false; LastDesc = &Desc->NextDesc, ++Desc);
-
-   // Add new description
    oldMap = Map.Data();
    map_ptrloc const descindex = NewDescription(Desc, List.DescriptionLanguage(), List.Description_md5(), *LastDesc);
    if (oldMap != Map.Data())
@@ -1403,3 +1422,14 @@ bool pkgCacheGenerator::MakeOnlyStatusCache(OpProgress *Progress,DynamicMMap **O
    return true;
 }
 									/*}}}*/
+// IsDuplicateDescription						/*{{{*/
+bool IsDuplicateDescription(pkgCache::DescIterator Desc,
+			    MD5SumValue const &CurMd5, std::string const &CurLang)
+{
+   for ( ; Desc.end() == false; ++Desc)
+      if (MD5SumValue(Desc.md5()) == CurMd5 && Desc.LanguageCode() == CurLang)
+	 return true;
+   return false;
+}
+									/*}}}*/
+
diff --git a/debian/changelog b/debian/changelog
index 1e5babe41..b7f199609 100644
--- a/debian/changelog
+++ b/debian/changelog
@@ -24,6 +24,7 @@ apt (0.8.16~exp7) UNRELEASEDexperimental; urgency=low
     - do not builtin languages only if none is forced (Closes: #643787)
   * apt-pkg/pkgcachegen.cc:
     - refactor MergeList by creating -Group, -Package and -Version specialist
+    - share description list between "same" versions (LP: #868977)
 
   [ Michael Vogt ]
   * apt-pkg/contrib/configuration.cc:
@@ -38,7 +39,7 @@ apt (0.8.16~exp7) UNRELEASEDexperimental; urgency=low
   * ftparchive/cachedb.cc:
     - fix buffersize in bytes2hex
 
- -- David Kalnischkies <kalnischkies@gmail.com>  Tue, 11 Oct 2011 18:32:13 +0200
+ -- David Kalnischkies <kalnischkies@gmail.com>  Tue, 11 Oct 2011 21:07:38 +0200
 
 apt (0.8.16~exp6) experimental; urgency=low
 
diff --git a/test/integration/test-bug-601016-description-translation b/test/integration/test-bug-601016-description-translation
index 44ab91900..03fddbfda 100755
--- a/test/integration/test-bug-601016-description-translation
+++ b/test/integration/test-bug-601016-description-translation
@@ -4,7 +4,7 @@ set -e
 TESTDIR=$(readlink -f $(dirname $0))
 . $TESTDIR/framework
 setupenvironment
-configarchitecture "i386"
+configarchitecture 'i386' 'amd64'
 
 # we need a valid locale here, otherwise the language configuration
 # will be overridden by LC_ALL=C
@@ -22,8 +22,22 @@ Size: 2140230
 MD5sum: 74769bfbcef9ebc4fa74f7a5271b9c08
 Description-md5: d41ee493aa9fcc6cbc9ce4eb7069959c"
 
+PACKAGESTANZA2='Package: apt
+Priority: important
+Section: admin
+Installed-Size: 5984
+Maintainer: APT Development Team <deity@lists.debian.org>
+Architecture: amd64
+Version: 0.8.7
+Filename: pool/main/a/apt/apt_0.8.7_amd64.deb
+Size: 2210342
+MD5sum: 4a869bfbdef9ebc9fa74f7a5271e8d1a
+Description-md5: d41ee493aa9fcc6cbc9ce4eb7069959c'
 
 echo "$PACKAGESTANZA
+Description: Advanced front-end for dpkg
+
+$PACKAGESTANZA2
 Description: Advanced front-end for dpkg" > aptarchive/Packages
 
 echo "Package: apt
@@ -53,6 +67,13 @@ Description-${LOCALE}: Mächtige Oberfläche für dpkg
  APT-Dselect-Methode. Beides sind einfache und sicherere Wege,
  um Pakete zu installieren und Upgrades durchzuführen.
 "
+LOCALESTANZA2="$PACKAGESTANZA2
+Description-${LOCALE}: Mächtige Oberfläche für dpkg
+ Das Paket bietet dem Nutzer technisch führende Methoden für den Zugriff
+ auf den dpkg-Paketmanager. Es beinhaltet das apt-get-Werkzeug und die
+ APT-Dselect-Methode. Beides sind einfache und sicherere Wege,
+ um Pakete zu installieren und Upgrades durchzuführen.
+"
 
 testrun() {
 	echo "Acquire::Languages { \"${LOCALE}\"; \"en\"; };" > rootdir/etc/apt/apt.conf.d/00languages
@@ -60,6 +81,8 @@ testrun() {
 	rm -rf rootdir/var/lib/apt/lists
 	setupaptarchive
 	testequal "$LOCALESTANZA" aptcache show apt -o Test=File-${LOCALE}
+	testequal "$LOCALESTANZA" aptcache show apt:i386 -o Test=File-${LOCALE}
+	testequal "$LOCALESTANZA2" aptcache show apt:amd64 -o Test=File-${LOCALE}
 	testequal "$NOLONGSTANZA" aptcache show apt -o Acquire::Languages="ww" -o Test=File-${LOCALE}
 	testequal "$LOCALESTANZA" aptcache show apt -o Acquire::Languages::="ww" -o Test=File-${LOCALE}
 	LC_ALL=C testequal "$ENGLISHSTANZA" aptcache show apt -o Test=File-${LOCALE}
@@ -73,6 +96,9 @@ testrun() {
 testrun
 
 echo "$PACKAGESTANZA
+Description: Advanced front-end for dpkg
+
+$PACKAGESTANZA2
 Description: Advanced front-end for dpkg" > aptarchive/Packages
 
 echo "Package: apt
@@ -88,5 +114,11 @@ Description-en: Advanced front-end for dpkg
  It provides the apt-get utility and APT dselect method that provides a
  simpler, safer way to install and upgrade packages.
 "
+ENGLISHSTANZA2="$PACKAGESTANZA2
+Description-en: Advanced front-end for dpkg
+ This is Debian's next generation front-end for the dpkg package manager.
+ It provides the apt-get utility and APT dselect method that provides a
+ simpler, safer way to install and upgrade packages.
+"
 
 testrun