From 0a4430109f3d1d3e757ec6d29b87fb74963ee2f2 Mon Sep 17 00:00:00 2001
From: Bill Allombert <ballombe@debian.org>
Date: Mon, 30 May 2011 20:42:12 +0000
Subject: [PATCH] popularity-contest: find package files lists by running dpkg
 -L by batches.

---
 debian/changelog   |  7 ++++++
 popularity-contest | 58 +++++++++++++++++++++++++++++++++++-----------
 2 files changed, 51 insertions(+), 14 deletions(-)

diff --git a/debian/changelog b/debian/changelog
index 98f64bb..6775c3d 100644
--- a/debian/changelog
+++ b/debian/changelog
@@ -1,3 +1,10 @@
+popularity-contest (1.53) UNRELEASED; urgency=low
+
+  * popularity-contest: 
+    - find package files lists by running dpkg -L by batches.  Closes: #622322
+
+ -- Bill Allombert <ballombe@debian.org>  Mon, 30 May 2011 17:04:20 +0200
+
 popularity-contest (1.52) unstable; urgency=low
 
   * Please note when reading stats that squeeze was released with 1.49.
diff --git a/popularity-contest b/popularity-contest
index 2375101..555ae0c 100755
--- a/popularity-contest
+++ b/popularity-contest
@@ -25,7 +25,7 @@
 use strict;
 use 5.6.0;
 
-my $dpkg_db="/var/lib/dpkg/info";
+my $dpkg_batch_size=32; #This is a time/space trade-off: higher = faster
 my $popcon_conf="/etc/popularity-contest.conf";
 
 my %opts=();
@@ -92,23 +92,14 @@ if (opendir(PROC, "/proc"))
   }
 }
 
-# Read dpkg database of installed packages
-open PACKAGES, "dpkg-query --show --showformat='\${status} \${package}\\n'|";
-while (<PACKAGES>)
+#process the file list of a package
+sub proc_pkg
 {
-  /^.*installed *(.+)$/ or next;
-  my $pkg=$1;
+  my ($pkg,@list)=@_;
   $popcon{$pkg}=[0,0,$pkg,"<NOFILES>"];
-  open FILES, "$dpkg_db/$pkg.list" or 
-    do { print STDERR "popcon: file $dpkg_db/$pkg.list is missing\n"; next; };
   my $bestatime = undef;
-  while (<FILES>)
+  for (@list)
   {
-    chop;
-    next unless (
-     ( m{/bin/|/sbin/|/lib/.+/|^/usr/games/|\.[ah]$|\.pm$|\.php$|^/boot/System\.map-}
-       || defined $mapped{$_} )
-       && -f $_);
     my($dev,$ino,$mode,$nlink,$uid,$gid,$rdev,$size,
                       $atime,$mtime,$ctime,$blksize,$blocks)
                           = stat;
@@ -138,9 +129,48 @@ while (<PACKAGES>)
       }
     }
   }
+}
+
+#process a list of packages.
+sub proc_pkgs
+{
+  my (@pkgs)=@_;
+  open FILES, "-|", "dpkg -L @pkgs";
+  my @list = ();
+  my $pkg = shift @pkgs;
+  while (<FILES>)
+  {
+    chop;
+    if ($_ eq "")
+    {
+      proc_pkg($pkg, @list);
+      $pkg = shift @pkgs;
+      @list = ();
+    }
+    next unless (
+     ( m{/bin/|/sbin/|/lib/.+/|^/usr/games/|\.[ah]$|\.pm$|\.php$|^/boot/System\.map-}
+       || defined $mapped{$_} )
+       && -f $_);
+    push @list, $_;
+  }
+  proc_pkg($pkg, @list);
   close FILES;
 }
 
+# Read dpkg database of installed packages
+open PACKAGES, "dpkg-query --show --showformat='\${status} \${package}\\n'|";
+my @pkglist = ();
+while (<PACKAGES>)
+{
+  /^.*installed *(.+)$/ or next;
+  push @pkglist, $1;
+  if (scalar @pkglist >= $dpkg_batch_size)
+  {
+    proc_pkgs(@pkglist);
+    @pkglist = ();
+  }
+}
+proc_pkgs(@pkglist) if (@pkglist);
 close PACKAGES;
 
 # We're not done yet.  Sort the output in reverse by atime, and