# description -- lintian check script -*- perl -*-

# Copyright (C) 1998 Christian Schwarz
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
#
# 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.  If not, you can find it on the World Wide
# Web at http://www.gnu.org/copyleft/gpl.html, or write to the Free
# Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston,
# MA 02110-1301, USA.

package Lintian::description;
use warnings;
use strict;

use Encode qw(decode);

use Lintian::Check qw(check_spelling check_spelling_picky);
use Lintian::Tags qw(tag);
use Util;

sub run {

my $pkg = shift;
my $type = shift;
my $info = shift;

my $tabs = 0;
my $lines = 0;
my $template = 0;
my $unindented_list = 0;
my $synopsis;
my $description;

# description?
my $full_description = $info->field('description');
unless (defined $full_description) {
    tag 'package-has-no-description', '';
    return 0;
}

$full_description =~ m/^([^\n]*)\n(.*)$/s;
($synopsis, $description) = ($1, $2);
unless (defined $synopsis) {
    $synopsis = $full_description;
    $description = '';
}

$description = '' unless defined($description);

if ($synopsis =~ m/^\s*$/) {
    tag 'description-synopsis-is-empty';
} else {
    if ($synopsis =~ m/^\Q$pkg\E\b/i) {
        tag 'description-starts-with-package-name';
    }
    if ($synopsis =~ m/^(an?|the)\s/i) {
        tag 'description-synopsis-starts-with-article';
    }
    if ($synopsis =~ m/(?<!etc)\.\s*$/i) {
        tag 'description-synopsis-might-not-be-phrased-properly';
    }
    if ($synopsis =~ m/\t/) {
        tag 'description-contains-tabs' unless $tabs++;
    }
    if ($synopsis =~ m/^missing\s*$/i) {
        tag 'description-is-debmake-template' unless $template++;
    } elsif ($synopsis =~ m/<insert up to 60 chars description>/) {
        tag 'description-is-dh_make-template' unless $template++;
    }

    # We have to decode into UTF-8 to get the right length for the length
    # check.  If the changelog uses a non-UTF-8 encoding, this will mangle it,
    # but it doesn't matter for the length check.
    if (length(decode('utf-8', $synopsis)) >= 80) {
        tag 'description-too-long';
    }
}

my $flagged_homepage;
foreach (split /\n/, $description) {
    next if m/^ \.\s*$/o;

    if ($lines == 0) {
        my $firstline = lc $_;
        my $lsyn = lc $synopsis;
        if ($firstline =~ /^\Q$lsyn\E$/) {
            tag 'description-synopsis-is-duplicated';
        } else {
            $firstline =~ s/[^a-zA-Z0-9]+//g;
            $lsyn =~ s/[^a-zA-Z0-9]+//g;
            if ($firstline eq $lsyn) {
                tag 'description-synopsis-is-duplicated';
            }
        }
    }

    $lines++;

    if (m/^ \.\s*\S/o) {
        tag 'description-contains-invalid-control-statement';
    } elsif (m/^ [\-\*]/o) {
        # Print it only the second time.  Just one is not enough to be sure that
        # it's a list, and after the second there's no need to repeat it.
        tag 'possible-unindented-list-in-extended-description' if $unindented_list++ == 2;
    }

    if (m/\t/o) {
        tag 'description-contains-tabs' unless $tabs++;
    }

    if (m,^\s*Homepage: <?https?://,i) {
        tag 'description-contains-homepage';
        $flagged_homepage = 1;
    }

    if (m,This description was automagically extracted from the module by dh-make-perl,i) {
        tag 'description-contains-dh-make-perl-template';
    }

    # Check for duplicated words.  We want to catch "this this." but not
    # "ITU-T T.81", so compare non-whitespace sequences rather than word
    # characters but allow punctuation at the end.
    #
    # We don't want to think ", ," or "a, a" is a duplicated word, so require
    # that a word start and end with a word character.
    #
    # We replace text that is quoted with ' "" '.  The assumption is that
    # quoted words are "okay" and blindly removing them causes false
    # positives with text like "'a' or 'b' or 'c'".
    my $stripped = $_;
    $stripped =~ s,(["'])(.*?)(\1), "" ,g;
    while ($stripped =~
           m%(?:\s|^)((\w(?:\S*\w)?)(\s+(\2))+)(?:[\).,?!:;\s]|\z)%i) {
        my $words = $1;
        $stripped =~ s/\Q$words//;
        tag 'description-contains-duplicated-word', $words;
    }

    my $first_person = $_;
    while ($first_person =~
           m/(?:^|\s)(I|[Mm]y|[Oo]urs?|mine|myself|me|us|[Ww]e)(?:$|\s)/) {
        my $word = $1;
        $first_person =~ s/\Q$word//;
        tag 'using-first-person-in-description', "line $lines: $word";
    }

    if ($lines == 1) {
        # checks for the first line of the extended description:
        if (m/^ \s/o) {
            tag 'description-starts-with-leading-spaces';
        }
        if (m/^\s*missing\s*$/oi) {
            tag 'description-is-debmake-template' unless $template++;
        } elsif (m/<insert long description, indented with spaces>/) {
            tag 'description-is-dh_make-template' unless $template++;
        }
    }

    if (length(decode('utf-8', $_)) > 80) {
        tag 'extended-description-line-too-long';
    }
}

if ($type ne 'udeb') {
    if ($lines == 0) {
        tag 'extended-description-is-empty';
    } elsif ($lines <= 2 and not $synopsis =~ /(dummy|transition)/i) {
        tag 'extended-description-is-probably-too-short';
    } elsif ($description =~ /^ \.\s*\n|\n \.\s*\n \.\s*\n|\n \.\s*\n?$/) {
        tag 'extended-description-contains-empty-paragraph';
    }
}

# Check for a package homepage in the description and no Homepage field.  This
# is less accurate and more of a guess than looking for the old Homepage:
# convention in the body.
unless ($info->field('homepage') or $flagged_homepage) {
    if ($description =~ /homepage|webpage|website|url|upstream|web\s+site
                         |home\s+page|further\s+information|more\s+info
                         |official\s+site|project\s+home/xi
        and $description =~ m,\b(https?://[a-z0-9][^>\s]+),i) {
        tag 'description-possibly-contains-homepage', $1;
    } elsif ($description =~ m,\b(https?://[a-z0-9][^>\s]+)>?\.?\s*\z,i) {
        tag 'description-possibly-contains-homepage', $1;
    }
}

if ($description) {
    check_spelling('spelling-error-in-description', $description,
                   undef, {$pkg => 1});
    check_spelling_picky('capitalization-error-in-description', $description);
}

}

1;

# Local Variables:
# indent-tabs-mode: nil
# cperl-indent-level: 4
# End:
# vim: syntax=perl sw=4 sts=4 sr et