Attempt to fix a bug in translate_why_unselected on x86_64 where sometimes ids

were reported instead of human-readable dependencies.

1 files changed, 9 insertions, 3 deletions

diff --git a/urpm.pm b/urpm.pm
index db8e272f..504eff54 100644
--- a/urpm.pm
+++ b/urpm.pm
@@ -3211,14 +3211,20 @@ sub unselected_packages {
 sub uniq { my %l; $l{$_} = 1 foreach @_; grep { delete $l{$_} } @_ }
 
 sub translate_why_unselected {
-    my (undef, $state, @l) = @_;
+    my ($urpm, $state, @l) = @_;
 
-    map { my $rb = $state->{rejected}{$_}{backtrack};
+    map {
+	my $rb = $state->{rejected}{$_}{backtrack};
 	my @froms = keys %{$rb->{closure} || {}};
 	my @unsatisfied = @{$rb->{unsatisfied} || []};
 	my $s = join ", ", (
 	    (map { N("due to missing %s", $_) } @froms),
-	    (map { N("due to unsatisfied %s", $_) } uniq @unsatisfied),
+	    (map { N("due to unsatisfied %s", $_) } uniq map {
+		    #- XXX in theory we shouldn't need this, dependencies (and not ids) should
+		    #- already be present in @unsatisfied. But with biarch packages this is
+		    #- not always the case.
+		    /\D/ ? $_ : scalar($urpm->{depslist}[$_]->fullname)
+		} @unsatisfied),
 	    $rb->{promote} && !$rb->{keep} ? N("trying to promote %s", join(", ", @{$rb->{promote}})) : @{[]},
 	    $rb->{keep} ? N("in order to keep %s", join(", ", @{$rb->{keep}})) : @{[]},
 	);