#!/usr/bin/env perl
#
# Author: petr.danecek@sanger
#

use strict;
use warnings;
use Carp;
use Vcf;

my $opts = parse_params();
read_data($opts);

exit;

#--------------------------------

sub error
{
    my (@msg) = @_;
    if ( scalar @msg ) { confess @msg; }
    die
        "Usage: vcf-query [OPTIONS] file.vcf.gz\n",
        "Options:\n",
        "   -c, --columns <file|list>           List of comma-separated column names or one column name per line in a file.\n",
        "   -f, --format <string>               The default is '%CHROM:%POS\\t%REF[\\t%SAMPLE=%GT]\\n'\n",
        "   -l, --list-columns                  List columns.\n",
        "   -r, --region chr:from-to            Retrieve the region. (Runs tabix.)\n",
        "       --use-old-method                Use old version of API, which is slow but more robust.\n",
        "   -h, -?, --help                      This help message.\n",
        "Expressions:\n",
        "   %CHROM          The CHROM column (similarly also other columns)\n",
        "   %GT             Translated genotype (e.g. C/A)\n",
        "   %GTR            Raw genotype (e.g. 0/1)\n",
        "   %INFO/TAG       Any tag in the INFO column\n",
        "   %LINE           Prints the whole line\n",
        "   %SAMPLE         Sample name\n",
        "   []              The brackets loop over all samples\n",
        "   %*<A><B>        All format fields printed as KEY<A>VALUE<B>\n",
        "Examples:\n",
        "   vcf-query file.vcf.gz 1:1000-2000 -c NA001,NA002,NA003\n",
        "   vcf-query file.vcf.gz -r 1:1000-2000 -f '%CHROM:%POS\\t%REF\\t%ALT[\\t%SAMPLE:%*=,]\\n'\n",
        "   vcf-query file.vcf.gz -f '[%GT\\t]%LINE\\n'\n",
        "   vcf-query file.vcf.gz -f '[%GT\\ ]%LINE\\n'\n",
        "   vcf-query file.vcf.gz -f '%CHROM\\_%POS\\t%INFO/DP\\t%FILTER\\n'\n",
        "\n";
}


sub parse_params
{
    my $opts = { columns=>'', format_string=>"%CHROM:%POS\t%REF[\t%SAMPLE=%GT]\n" };
    while (defined(my $arg=shift(@ARGV)))
    {
        if (                 $arg eq '--use-old-method' ) { $$opts{use_old_method}=1; next }
        if ( $arg eq '-f' || $arg eq '--format' ) { $$opts{format_string}=shift(@ARGV); next }
        if ( $arg eq '-c' || $arg eq '--columns' ) { $$opts{columns}=shift(@ARGV); next }
        if ( $arg eq '-l' || $arg eq '--list-columns' ) { $$opts{list_columns}=1; next }
        if ( $arg eq '-r' || $arg eq '--region' ) { $$opts{region}=shift(@ARGV); next }
        if ( -e $arg or $arg=~m{^(?:ftp|http)://} ) { $$opts{file}=$arg; next; }
        if ( $arg eq '-?' || $arg eq '-h' || $arg eq '--help' ) { error(); }
        if ( !exists($$opts{region}) && exists($$opts{file}) && ($arg=~/^[^:]+:[0-9,]+-[0-9,]+$/ or $arg=~/^[^\:]+$/) ) { $$opts{region}=$arg; next; }
        error("Unknown parameter or non-existent file \"$arg\". Run -h for help.\n");
    }
    if ( !exists($$opts{file}) && exists($$opts{region}) ) { error("The region cannot be used when streaming the file.\n"); }
    if ( exists($$opts{columns}) && -e $$opts{columns} )
    {
        my @cols;
        open(my $fh,'<',$$opts{columns}) or error("$$opts{columns}: $!");
        while (my $line=<$fh>)
        {
            if ( $line=~/^\s*$/ ) { next; }
            $line =~ s/^\s*//;
            $line =~ s/\s*$//;
            push @cols, $line;
        }
        close($fh);
        $$opts{columns} = join(',', @cols);
    }
    return $opts;
}

sub parse_format_string
{
    my ($str,$hash) = @_;
    my (@arr,%idx,$join1,$join2);
    $str =~ s/\\n/\n/g;
    $str =~ s/\\t/\t/g;
    while ($str)
    {
        if ( !($str=~/%/) )
        {
            push @arr,$str;
            last;
        }

        my $before = $`;
        $str = $';

        my $match;
        if ( $str=~/^[*](.)(.)/ )
        {
            $match = '*'; $join1=$1; $join2=$2;
        }
        elsif ( $str=~m{([A-Za-z0-9/_]+)} )
        {
            $match = $1;
        }
        else { error("FIXME: $str"); }

        if ( defined $before && $before ne '' ) { push @arr,$before; }
        push @arr,'.';      # If the tag is not present in the VCF, a missing value ('.') will be printed instead.
        if ( exists($idx{$match}) ) 
        { 
            warn("The tag \"$match\" given multiple times, only the last occurance will be used\n"); 
        }
        $idx{$match} = $#arr;
        $str = $';
    }
    for (my $i=0; $i<@arr; $i++)
    {
        $arr[$i] =~ s/\\{1}//g;
    }
    $$hash{format} = \@arr;
    $$hash{idx}    = \%idx;
    $$hash{join1}  = $join1;
    $$hash{join2}  = $join2;
}

sub parse_format
{
    my ($opts,$cols) = @_;

    $$opts{before} = {};
    $$opts{repeat} = {};
    $$opts{after}  = {};

    my ($before,$repeat,$after);

    my $str = $$opts{format_string};
    $before = $str;

    if ( $str=~/\[([^\]]+)\]/ )
    {
        $before = $`;
        $repeat = $1;
        $after  = $';
    }
    if ( $before ) { parse_format_string($before,$$opts{before}); }
    if ( $repeat ) { parse_format_string($repeat,$$opts{repeat}); }
    if ( $after ) { parse_format_string($after,$$opts{after}); }
}

sub copy_array
{
    my ($arr) = @_;
    my @out;
    for my $item (@$arr) { push @out,$item; }
    return @out;
}

sub get_columns
{
    my ($vcf) = @_;
    my @cols = ();
    my $ncols = @{$$vcf{columns}};
    for (my $i=9; $i<$ncols; $i++)
    {
        push @cols, $$vcf{columns}[$i];
    }
    return \@cols;
}

sub get_sample_idxs
{
    my ($vcf,@samples) = @_;
    my @idxs;
    for my $sample (@samples)
    {
        if ( !exists($$vcf{has_column}{$sample}) ) { error("No such sample: [$sample]\n"); }
        push @idxs, $$vcf{has_column}{$sample} - 1;
    }
    return @idxs;
}

sub list_columns
{
    my ($opts) = @_;
    my $cols = get_columns($$opts{vcf});
    for my $col (@$cols) { print "$col\n"; }
}

sub read_data
{
    my ($opts) = @_;

    if ( exists($$opts{use_old_method}) )
    {
        read_data_slow_hash($opts);
        return;
    }

    my %args = ( print_header=>1 );
    if ( $$opts{region} ) { $args{region} = $$opts{region}; }
    if ( exists($$opts{file}) ) { $args{file} = $$opts{file}; }
    else { $args{fh} = \*STDIN; }

    my $vcf = Vcf->new(%args);
    $$opts{vcf} = $vcf;
    $vcf->parse_header();

    if ( $$opts{list_columns} ) { list_columns($opts); exit; }

    my @cols = split(/,/,$$opts{columns});
    if ( !@cols ) { @cols = @{get_columns($$opts{vcf})}; }
    my @sample_idxs = get_sample_idxs($$opts{vcf},@cols);

    # The hash opts will be filled with the keys 'before','repeat','after' with formatting information
    parse_format($opts);

    while (my $line=$vcf->next_line())
    {
        my $x = $vcf->next_data_array($line);

        # Fill everything what comes before the repeat []
        if ( $$opts{before} )
        {
            my (@out) = copy_array($$opts{before}{format});
            while (my ($fieldname,$idx) = each %{$$opts{before}{idx}})
            {
                if ( $fieldname eq 'LINE' ) { chomp($line); $out[$idx] = $line; }
                elsif ( exists($$vcf{has_column}{$fieldname}) ) { $out[$idx] = $$x[$$vcf{has_column}{$fieldname}-1]; }
                elsif ( substr($fieldname,0,5) eq 'INFO/' )
                {
                    $out[$idx] = $vcf->get_info_field($$x[7],substr($fieldname,5));
                }
            }
            for (my $i=0; $i<@out; $i++) { if (!defined($out[$i])) { $out[$i]='.'; } }
            print join('',@out);
        }

        # Fill the repeaty stuff (the sample columns)
        if ( $$opts{repeat} )
        {
            my @repeats;
            for my $sample_idx (@sample_idxs) { push @repeats, [ copy_array($$opts{repeat}{format}) ]; }

            my @alt;
            if ( exists($$opts{repeat}{idx}{GT}) )
            {
                @alt = split(/,/,$$x[4]);
            }
            while (my ($fieldname,$idx) = each %{$$opts{repeat}{idx}})
            {
                if ( $fieldname eq '*' )
                {
                    my $sep1 = $$opts{repeat}{join1};
                    my $sep2 = $$opts{repeat}{join2};
                    my @fmt = split(/:/,$$x[8]);
                    for (my $i=0; $i<@sample_idxs; $i++)
                    {
                        my $sample_idx = $sample_idxs[$i];
                        my @tmp;
                        my $j = 0;
                        for my $value (split(/:/,$$x[$sample_idx])) 
                        { 
                            push @tmp, $fmt[$j++].$sep1.$value; 
                        }
                        $repeats[$i][$idx] = join($sep2,@tmp);
                    }
                    next;
                }

                my $fmt_idx = $vcf->get_tag_index($$x[8],$fieldname eq 'GTR' ? 'GT' : $fieldname,':');
                for (my $i=0; $i<@sample_idxs; $i++)
                {
                    my $sample_idx = $sample_idxs[$i];
                    if ( $fmt_idx!=-1 )
                    {
                        my $value = $vcf->get_field($$x[$sample_idx],$fmt_idx);
                        if ( $fieldname eq 'GT' )
                        {
                            $value = $vcf->decode_genotype($$x[3],\@alt,$value);
                        }
                        $repeats[$i][$idx] = $value;
                    }
                }
            }
            if ( exists($$opts{repeat}{idx}{SAMPLE}) )
            {
                my $idx = $$opts{repeat}{idx}{SAMPLE};
                for (my $i=0; $i<@cols; $i++) { $repeats[$i][$idx] = $cols[$i] }
            }

            for my $repeat (@repeats)
            {
                for (my $i=0; $i<@$repeat; $i++) { if (!defined($$repeat[$i])) { $$repeat[$i]='.'; } }
                print join('',@$repeat);
            }
        }

        # Fill everything what comes after the repeat ([])
        if ( $$opts{after} )
        {
            my (@out) = copy_array($$opts{after}{format});
            while (my ($fieldname,$idx) = each %{$$opts{after}{idx}})
            {
                if ( $fieldname eq 'LINE' ) { chomp($line); $out[$idx] = $line; }
                elsif ( exists($$vcf{has_column}{$fieldname}) ) { $out[$idx] = $$x[$$vcf{has_column}{$fieldname}-1]; }
                elsif ( substr($fieldname,0,5) eq 'INFO/' )
                {
                    $out[$idx] = $vcf->get_info_field($$x[7],substr($fieldname,5));
                }
            }
            for (my $i=0; $i<@out; $i++) { if (!defined($out[$i])) { $out[$i]='.'; } }
            print join('',@out);
        }
    }
}


sub read_data_slow_hash
{
    my ($opts) = @_;

    my %args = ( print_header=>1 );
    if ( $$opts{region} ) { $args{region} = $$opts{region}; }
    if ( exists($$opts{file}) ) { $args{file} = $$opts{file}; }
    else { $args{fh} = \*STDIN; }

    my $vcf = Vcf->new(%args);
    $$opts{vcf} = $vcf;
    $vcf->parse_header();

    if ( $$opts{list_columns} ) { list_columns($opts); exit; }

    my @cols = split(/,/,$$opts{columns});
    if ( !@cols ) { @cols = @{get_columns($$opts{vcf})}; }

    # The hash opts will be filled with the keys 'before','repeat','after' with formatting information
    parse_format($opts);

    while (my $line=$vcf->next_line())
    {
        my $x=$vcf->next_data_hash($line);

        # Fill everything what comes before the repeat []
        # Code repetition and not very nice, should be changed at some point...
        if ( $$opts{before} )
        {
            my (@out) = copy_array($$opts{before}{format});
            while (my ($colname,$idx) = each %{$$opts{before}{idx}})
            {
                if ( $colname eq 'LINE' ) { chomp($line); $out[$idx] = $line; next; }
                if ( $colname eq 'ALT' ) { $out[$idx] = join(',',@{$$x{ALT}}); next; }
                if ( $colname eq 'FILTER' ) { $out[$idx] = join(';',@{$$x{FILTER}}); next; }
                if ( $colname=~m{INFO/(.+)} ) 
                { 
                    if ( exists($$x{INFO}{$1}) && !defined($$x{INFO}{$1}) )
                    {
                        # It is a flag
                        $out[$idx] = 'True';
                    }
                    else
                    {
                        $out[$idx] = $$x{INFO}{$1}; 
                    }
                    next; 
                }
                if ( exists($$x{$colname}) ) { $out[$idx] = $$x{$colname}; }
            }
            for (my $i=0; $i<@out; $i++) { if (!defined($out[$i])) { $out[$i]='.'; } }
            print join('',@out);
        }

        # Fill the repeaty stuff (the sample columns)
        if ( $$opts{repeat} )
        {
            for my $col (@cols)
            {
                my ($alleles,$seps,$is_phased,$is_empty) = $vcf->parse_haplotype($x,$col);
                my (@out) = copy_array($$opts{repeat}{format});
                while (my ($colname,$idx) = each %{$$opts{repeat}{idx}})
                {
                    if ( exists($$x{gtypes}{$col}{$colname}) ) { $out[$idx] = $$x{gtypes}{$col}{$colname}; }
                    elsif ( exists($$x{$colname}) ) { $out[$idx] = $$x{$colname}; }
                }
                if ( exists($$opts{repeat}{idx}{SAMPLE}) ) { $out[$$opts{repeat}{idx}{SAMPLE}] = $col; }
                if ( exists($$opts{repeat}{idx}{GTR}) ) 
                { 
                    $out[$$opts{repeat}{idx}{GTR}] = $$x{gtypes}{$col}{GT};
                }
                if ( exists($$opts{repeat}{idx}{GT}) ) 
                { 
                    my $tmp = $$alleles[0];
                    for (my $i=0; $i<@$seps; $i++) { $tmp .= $$seps[$i].$$alleles[$i+1]; }
                    $out[$$opts{repeat}{idx}{GT}] = $tmp;
                }
                if ( exists($$opts{repeat}{idx}{'*'}) )
                {
                    my $sep1 = $$opts{repeat}{join1};
                    my $sep2 = $$opts{repeat}{join2};
                    my @tmp;
                    while (my ($key,$value)=each(%{$$x{gtypes}{$col}}))
                    {
                        if ( $key eq 'GT' )
                        { 
                            $value = $$alleles[0];
                            for (my $i=0; $i<@$seps; $i++) { $value .= $$seps[$i].$$alleles[$i+1]; }
                        }
                        push @tmp, $key.$sep1.$value;
                    }
                    my $idx = $$opts{repeat}{idx}{'*'};
                    $out[$idx] = join($sep2,@tmp);
                }
                for (my $i=0; $i<@out; $i++) { if (!defined($out[$i])) { $out[$i]='.'; } }
                print join('',@out);
            }
        }

        # Fill everything what comes after the repeat ([])
        if ( $$opts{after} )
        {
            my (@out) = copy_array($$opts{after}{format});
            while (my ($colname,$idx) = each %{$$opts{after}{idx}})
            {
                if ( $colname eq 'LINE' ) { chomp($line); $out[$idx] = $line; next; }
                if ( $colname eq 'ALT' ) { $out[$idx] = join(',',@{$$x{ALT}}); next; }
                if ( $colname eq 'FILTER' ) { $out[$idx] = join(';',@{$$x{FILTER}}); next; }
                if ( $colname=~m{INFO/(.+)} ) 
                { 
                    if ( exists($$x{INFO}{$1}) && !defined($$x{INFO}{$1}) )
                    {
                        # It is a flag
                        $out[$idx] = 'True';
                    }
                    else
                    {
                        $out[$idx] = $$x{INFO}{$1}; 
                    }
                    next; 
                }
                if ( exists($$x{$colname}) ) { $out[$idx] = $$x{$colname}; }
            }
            for (my $i=0; $i<@out; $i++) { if (!defined($out[$i])) { $out[$i]='.'; } }
            print join('',@out);
        }
    }
}