#!/usr/bin/perl -w
#
# Copyright (c) 2014 Adrian Schroeter, SUSE LLC
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License version 2 or 3 as
# published by the Free Software Foundation.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program (see the file COPYING); if not, write to the
# Free Software Foundation, Inc.,
# 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA
#
################################################################

BEGIN {
  my ($wd) = $0 =~ m-(.*)/- ;
  $wd ||= '.';
  unshift @INC,  "$wd/build";
  unshift @INC,  "$wd";
}

use strict;

use POSIX;
use Data::Dumper;
use Getopt::Long;
use Date::Parse;

sub echo_help {
    print "\n
The SUSE changes merge tool
===========================

openSUSE package sources contain a .changes file providing a full changelog.
A stripped version of that gets attached to rpm %changes to avoid a too large rpm db.
But the full version is kept with the package sources.

The classic diff3 merge approach often fails on these files, so this
merge tool is reading the entire files and sorts the entries according
to it's date. If entries with same date do differ it fails. The classic
diff3 merge and manual conflict resolution is the only way then.

The tool takes any amount of files and is merging them into one. The first file
must be the common ancestor of the other files.

";
}

#
# Argument parsing
#
if (@ARGV < 1) {
  echo_help();
  exit(1);
}

my @files;
my $force;
my $verbose;

while (@ARGV) {
  my $arg = shift @ARGV;
  if ($arg eq "--help") {
    echo_help();
    exit(0);
  }

  if ($arg eq "--verbose") {
    $verbose = 1;
  } elsif ($arg eq "--force") {
    $force = 1;
  } elsif ($arg =~ /^-/) {
    die("Unknown switch $arg");
  } else {
    push @files, $arg;
  }
}

die("Give at least one file") if @files < 1;

# init
my $seperator = "-------------------------------------------------------------------";
my %entries;

# utils

sub time2mystr {
  my ($time) = @_;
  my @lt = gmtime($time);
  # ctime(3) format output
  return strftime("%a %b %e %H:%M:%S UTC %Y", @lt);
}

sub findsim {
  my ($ent, $allents) = @_;
  return undef unless $ent->{'fileno'};
  my %sim;
  my @in = grep {!$_->{'fileno'}} @$allents;
  for my $ent2 (@in) {
    return $ent2 if $ent2->{'text'} eq $ent->{'text'};
  }
  for my $ent2 (@in) {
    my @w = split(' ', $ent2->{'text'});
    my @wc = grep {$ent->{'text'} =~ /\Q$_\E/} @w;
    $sim{$ent2} = @wc ? @w / @wc : 0;
  }
  my @sorted = sort {$sim{$a} <=> $sim{$b}} @in;
  return $sorted[-1] if @sorted && $sorted[-1] > .75;
  return undef;
}

sub decide {
  my ($ent, $nfiles) = @_;
  return $ent unless $ent->{'sim'};
  my @sim = @{$ent->{'sim'}};
  my @changed = grep {$_->{'text'} ne $ent->{'text'}} @sim;
  if (!@changed) {
    return $sim[-1] if @sim == $nfiles - 1;
    return undef if @sim == $nfiles - 2;
  }
  if (@changed && @sim == $nfiles - 1) {
    return $changed[-1] unless grep {$_->{'text'} ne $changed[-1]->{'text'}} @changed;
  }
  die("Conflicting entries for $ent->{time}\n");
}

sub setentry {
  my ($time, $timestr, $email, $text, $fileno) = @_;
  my $ent = {'time' => $timestr, 'email' => $email, 'text' => $text, 'fileno' => $fileno};
  if ($entries{$time}) {
    my $siment = findsim($ent, $entries{$time});
    if ($siment) {
      push @{$siment->{'sim'}}, $ent;
      return;
    }
  }
  push @{$entries{$time}}, $ent;
}

# read all files into a hash
my $fileno = 0;
my $nfiles = @files;
while (@files) {
  my $file = shift @files;
  local *F;
  open(F, '<', $file) || die("Unable to open $file");
  my @lines = <F>;
  close F;
  print "Read file $file\n" if $verbose;

  my $init;
  my $email;
  my $time;
  my $timestr;
  my $text = ""; 
  my $cycle = 0;
  foreach my $line (@lines) {
    chomp($line);
    if (!$init) {
      die("no ---- seperator in first line\n") unless $line eq $seperator;
      $init = 1;
      next;
    }

    if (!$time) {
      ($timestr, $email) = split(' - ', $line, 2);
      $time = str2time($timestr);
      die("unable to parse time $line\n") unless $time;
      die("unable to find email in $line\n") unless $email;
      print "Read ".time2mystr($time)."($time) for $line\n" if $verbose;
      next;
    }
    
    if ($line eq $seperator) {
      my @lt = gmtime($time);
      # check for the special case, we had many entries at 00:00:00 on same day in the past ...
      # ignoring the hour due to timezone issues, but do not accept it anymore for current entries
      # we take this as one blob.
      # Accept this only for entries in 2006 and before with 00 minutes and 00 seconds
      if ($lt[5] > 106 || $lt[1] != 0 || $lt[0] != 0) {
        setentry($time, $timestr, $email, $text, $fileno);
        $text = ""; 
        $time = undef;
        $email = undef;
        next;
      }
    }

    # must be text
    $text .= "$line\n";
  }
  # last entry
  setentry($time, $timestr, $email, $text, $fileno);
  $fileno++;
}

print "Merged ouput:\n===========\n" if $verbose;

# output the hash
for my $time (sort {$b <=> $a} keys %entries) {
  my %seen;
  for my $ent (@{$entries{$time}}) {
    $ent = decide($ent, $nfiles);
    next unless $ent && $ent->{'fileno'};	# ignore old stuff
    next if $seen{$ent->{text}};
    print "$seperator\n$ent->{time} - $ent->{email}\n$ent->{text}";
    $seen{$ent->{text}} = 1;
  }
}