Michael Tautschnig <m...@debian.org> writes:
> I've just checked and I think both the file one can download from the
> bugreport and my local copy here _do_ have a newline, but my methods for
> checking that might not be as effective. Anyway, I've attached a freshly
> generated index file, this time gzipped to make sure my mail agent
> doesnt harm it and I've also BCC'ed Russ to avoid any extra loops.
Yup. There's a trailing newline -- that wasn't the problem.
I'm a little disturbed by the degree to which the spacing indentation
changes in your index. That in and of itself shouldn't cause Lintian any
problems, but I don't understand why it's happening. It doesn't do this
for me when I run tar -tvf on a source tarball, and the manipulations that
Lintian does shouldn't change the spacing of the columns. I'm not sure if
that's perhaps a sign of some other issue or just nothing.
I *think* I might have fixed your problem, though. Could you try the
following patch against your /usr/share/lintian directory and see if it
makes the problem go away? (Although if this did fix it, I don't know why
the problem wasn't reproducible.)
--- a/unpack/unpack-srcpkg-l1
+++ b/unpack/unpack-srcpkg-l1
@@ -126,7 +126,7 @@ my $prefix;
for my $line (@index) {
my ($file) = ($line =~ /^(?:\S+\s+){5}(.*)/);
$file =~ s,^\./+,,;
- my ($dir) = ($file =~ m,([^/]+),);
+ my ($dir) = ($file =~ m,^([^/]+),);
if (defined($dir) and $dir eq $file and not $line =~ /^d/) {
$prefix = '';
} elsif (defined $dir) {
@@ -141,7 +141,7 @@ for my $line (@index) {
}
if ($prefix) {
@index = map {
- s,^((?:\S+\s+){5})(?:\./+)?\Q$prefix\E(?:/+|\z),$1,;
+ s,^((?:\S+\s+){5})(?:\./+)?\Q$prefix\E(?:/+|\Z),$1,;
if (/^(?:\S+\s+){5}\S+/) {
$_;
} else {
--
Russ Allbery (r...@stanford.edu) <http://www.eyrie.org/~eagle/>
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
