# File: FuncBuiltins.pm
#
# Purpose: Registers the basic built-in Functions

# SPDX-FileCopyrightText: 2020-2023 Pragmatic Software <pragma78@gmail.com>
# SPDX-License-Identifier: MIT

package PBot::Plugin::FuncBuiltins;
use parent 'PBot::Plugin::Base';

use PBot::Imports;

use PBot::Core::Utils::Indefinite;

use Lingua::EN::Tagger;
use URI::Escape qw/uri_escape_utf8/;

sub initialize($self, %conf) {
    $self->{pbot}->{functions}->register(
        'title',
        {
            desc   => 'Title-cases text',
            usage  => 'title <text>',
            subref => sub { $self->func_title(@_) }
        }
    );
    $self->{pbot}->{functions}->register(
        'ucfirst',
        {
            desc   => 'Uppercases first character',
            usage  => 'ucfirst <text>',
            subref => sub { $self->func_ucfirst(@_) }
        }
    );
    $self->{pbot}->{functions}->register(
        'uc',
        {
            desc   => 'Uppercases all characters',
            usage  => 'uc <text>',
            subref => sub { $self->func_uc(@_) }
        }
    );
    $self->{pbot}->{functions}->register(
        'lc',
        {
            desc   => 'Lowercases all characters',
            usage  => 'lc <text>',
            subref => sub { $self->func_lc(@_) }
        }
    );
    $self->{pbot}->{functions}->register(
        'unquote',
        {
            desc   => 'removes unescaped surrounding quotes and strips escapes from escaped quotes',
            usage  => 'unquote <text>',
            subref => sub { $self->func_unquote(@_) }
        }
    );
    $self->{pbot}->{functions}->register(
        'shquote',
        {
            desc   => 'quotes text for sh invocation',
            usage  => 'shquote <text>',
            subref => sub { $self->func_shquote(@_) }
        }
    );
    $self->{pbot}->{functions}->register(
        'quotemeta',
        {
            desc   => 'escapes/quotes metacharacters',
            usage  => 'quotemeta <text>',
            subref => sub { $self->func_quotemeta(@_) }
        }
    );
    $self->{pbot}->{functions}->register(
        'uri_escape',
        {
            desc   => 'percent-encode unsafe URI characters',
            usage  => 'uri_escape <text>',
            subref => sub { $self->func_uri_escape(@_) }
        }
    );
    $self->{pbot}->{functions}->register(
        'ana',
        {
            desc   => 'fix-up a/an article at front of text',
            usage  => 'ana <text>',
            subref => sub { $self->func_ana(@_) }
        }
    );
    $self->{pbot}->{functions}->register(
        'maybe-the',
        {
            desc   => 'prepend "the" in front of text depending on the part-of-speech of the first word in text',
            usage  => 'maybe-the <text>',
            subref => sub { $self->func_maybe_the(@_) }
        }
    );

    $self->{tagger} = Lingua::EN::Tagger->new;
}

sub unload($self) {
    $self->{pbot}->{functions}->unregister('title');
    $self->{pbot}->{functions}->unregister('ucfirst');
    $self->{pbot}->{functions}->unregister('uc');
    $self->{pbot}->{functions}->unregister('lc');
    $self->{pbot}->{functions}->unregister('unquote');
    $self->{pbot}->{functions}->unregister('shquote');
    $self->{pbot}->{functions}->unregister('quotemeta');
    $self->{pbot}->{functions}->unregister('uri_escape');
    $self->{pbot}->{functions}->unregister('ana');
    $self->{pbot}->{functions}->unregister('maybe-the');
}

sub func_unquote($self, @rest) {
    my $text = "@rest";
    $text =~ s/^"(.*?)(?<!\\)"$/$1/ || $text =~ s/^'(.*?)(?<!\\)'$/$1/;
    $text =~ s/(?<!\\)\\'/'/g;
    $text =~ s/(?<!\\)\\"/"/g;
    return $text;
}

sub func_title($self, @rest) {
    my $text = "@rest";
    $text = ucfirst lc $text;
    $text =~ s/ (\w)/' ' . uc $1/ge;
    return $text;
}

sub func_ucfirst($self, @rest) {
    my $text = "@rest";

    my ($word) = $text =~ m/^\s*([^',.;: ]+)/;

    # don't ucfirst on nicks
    if ($self->{pbot}->{nicklist}->is_present_any_channel($word)) {
        return $text;
    }

    return ucfirst $text;
}

sub func_uc($self, @rest) {
    my $text = "@rest";
    return uc $text;
}

sub func_lc($self, @rest) {
    my $text = "@rest";
    return lc $text;
}

sub func_shquote($self, @rest) {
    my $text = "@rest";
    $text =~ s/'/'"'"'/g;
    return "'$text'";
}

sub func_quotemeta($self, @rest) {
    my $text = "@rest";
    return quotemeta $text;
}

sub func_uri_escape($self, @rest) {
    my $text = "@rest";
    return uri_escape_utf8($text);
}

sub func_ana($self, @rest) {
    my $text = "@rest";

    if ($text =~ s/\b(an?)(\s+)//i) {
        my ($article, $spaces) = ($1, $2);
        my $fixed_article = select_indefinite_article $text;

        if ($article eq 'AN') {
            $fixed_article = uc $fixed_article;
        } elsif ($article eq 'An' or $article eq 'A') {
            $fixed_article = ucfirst $fixed_article;
        }

        $text = $fixed_article . $spaces . $text;
    }

    return $text;
}

sub func_maybe_the($self, @rest) {
    my $text = "@rest";

    my ($word) = $text =~ m/^\s*([^',.;: ]+)/;

    # don't prepend "the" if a proper-noun nick follows
    if ($self->{pbot}->{nicklist}->is_present_any_channel($word)) {
        return $text;
    }

    # special-case some indefinite nouns that Lingua::EN::Tagger treats as plain nouns
    if ($word =~ m/(some|any|every|no)(thing|one|body|how|way|where|when|time|place)/i) {
        return $text;
    }

    my $tagged = $self->{tagger}->add_tags($word);

    if ($tagged !~ m/^\s*<(?:det|prps?|cd|in|nnp|to|rb|wdt|rbr|jjr)>/) {
        $text = "the $text";
    }

    return $text;
}

1;