From 2c4302dd7a437c3153d6de704fb5ddec004ea308 Mon Sep 17 00:00:00 2001
From: Eelco Dolstra <e.dolstra@tudelft.nl>
Date: Mon, 28 Feb 2005 14:12:06 +0000
Subject: * Added a disambiguation heuristic: if two components have the same  
 name but differ to much in sice (by more than a factor of 3), then   never
 generate a patch.

---
 scripts/Makefile.am            |  2 +-
 scripts/generate-patches.pl.in | 26 +++++++++++++++++++++++---
 2 files changed, 24 insertions(+), 4 deletions(-)

diff --git a/scripts/Makefile.am b/scripts/Makefile.am
index 20e42e74c..8f42095b4 100644
--- a/scripts/Makefile.am
+++ b/scripts/Makefile.am
@@ -2,7 +2,7 @@ bin_SCRIPTS = nix-collect-garbage \
   nix-pull nix-push nix-prefetch-url \
   nix-install-package nix-channel nix-build
 
-noinst_SCRIPTS = nix-profile.sh
+noinst_SCRIPTS = nix-profile.sh generate-patches.pl
 
 nix-pull nix-push: readmanifest.pm download-using-manifests.pl
 
diff --git a/scripts/generate-patches.pl.in b/scripts/generate-patches.pl.in
index ef4ddcda4..87334b5fc 100755
--- a/scripts/generate-patches.pl.in
+++ b/scripts/generate-patches.pl.in
@@ -134,11 +134,11 @@ print "creating patches...\n";
 foreach my $p (keys %dstOutPaths) {
 
     # If exactly the same path already exists in the source, skip it.
-    next if defined $srcOutPaths{$p};
+# !!!    next if defined $srcOutPaths{$p};
     
-#    print "  $p\n";
+    print "  $p\n";
 
-    # If not, then we should find the path in the source that is
+    # If not, then we should find the paths in the source that are
     # `most' likely to be present on a system that wants to install
     # this path.
 
@@ -153,6 +153,22 @@ foreach my $p (keys %dstOutPaths) {
     foreach my $q (keys %srcOutPaths) {
         (my $name2, my $version2) = getNameVersion $q;
         if ($name eq $name2) {
+
+            # If the sizes differ to much, then skip.  This
+            # disambiguates between, e.g., a real component and a
+            # wrapper component (cf. Firefox in Nixpkgs).
+            my $srcSize = @{$srcNarFiles{$q}}[0]->{size};
+            my $dstSize = @{$dstNarFiles{$p}}[0]->{size};
+            my $ratio = $srcSize / $dstSize;
+            $ratio = 1 / $ratio if $ratio < 1;
+            print "  $srcSize $dstSize $ratio $q\n";
+
+            if ($ratio >= 3) {
+                print "  SKIPPING $q due to size ratio $ratio\n";
+            }
+
+            # If there are multiple matching names, include the ones
+            # with the closest version numbers.
             my $dist = versionDiff $version, $version2;
             if ($dist > $minDist) {
                 $minDist = $dist;
@@ -174,6 +190,8 @@ foreach my $p (keys %dstOutPaths) {
         # Generate a patch between $closest and $p.
         print "  $p <- $closest\n";
 
+        next;
+
         # If the patch already exists, skip it.
         if (containsPatch(\%srcPatches, $p, $closest) ||
             containsPatch(\%dstPatches, $p, $closest))
@@ -242,6 +260,8 @@ foreach my $p (keys %dstOutPaths) {
     }
 }
 
+exit 0; # !!!
+
 
 # Add in any potentially useful patches in the source (namely, those
 # patches that produce either paths in the destination or paths that
-- 
cgit v1.2.3