hpr-tools/Show_Submission/parse_JSON

#!/usr/bin/env perl
#===============================================================================
#
#         FILE: parse_JSON
#
#        USAGE: ./parse_JSON [-help] [-manpage] -episode=N -infile=FILE
#                       [-shownotes[=FILE]] [-format=FILE] [-release=FILE]
#                       [-assets=FILE] [-debug=N] [-[no]silent] [-[no]test]
#
#  DESCRIPTION: Script to parse the JSON now being generated for HPR
#               shows. Replaces the original 'parse_shownotes' and is a fair
#               bit simpler because the parsing of the original PHP dump isn't
#               needed. However, it has grown in complexity as features for
#               handling assets have been added. Downstream scripts now need
#               information about pictures and assets, and all archive files
#               need to be unpacked and placed in a directory for uploading to
#               the HPR server.
#
#      OPTIONS: ---
# REQUIREMENTS: ---
#         BUGS: ---
#        NOTES: ---
#       AUTHOR: Dave Morriss (djm), Dave.Morriss@gmail.com
#      VERSION: 0.0.17
#      CREATED: 2020-11-28 10:52:02
#     REVISION: 2024-10-04 18:37:29
#
#===============================================================================

use v5.36;
use utf8;
use feature qw{ postderef say signatures state try };
no warnings
    qw{ experimental::postderef experimental::signatures experimental::try };

use Getopt::Long;
use Pod::Usage;

use Term::ANSIColor;

use HTML::Parser ();
use HTML::LinkExtor;
use Text::CSV;
use Text::SpellChecker;

use JSON;

use List::MoreUtils qw{any};
use Archive::Any;
use MIME::Types;
use IO::Compress::Zip qw(:all);
use File::Copy;
use File::Spec::Functions;
use Path::Class;

use Log::Handler;

use Data::Dumper;

#
# Version number (manually incremented)
#
our $VERSION = '0.0.17';

#
# Script and directory names
#
( my $PROG = $0 ) =~ s|.*/||mx;
( my $DIR  = $0 ) =~ s|/?[^/]*$||mx;
$DIR = '.' unless $DIR;

#-------------------------------------------------------------------------------
# Declarations
#-------------------------------------------------------------------------------
#
# Constants and other declarations
#
my $basedir = "$ENV{HOME}/HPR/Show_Submission";
my $logdir  = "$basedir/logs";
my $logfile = "$logdir/${PROG}.log";

my ( $snlevel, $showno, $summarylength, $notelength );

our $MARKUP_DETECTED = 0;

#
# Maximum and minimum number of characters (bytes)
#
my $MAXNOTELEN = 4000;
my $MINNOTELEN = 10;

#
# Printing: general output format
#
my $ofmt = "%-17s %s\n";

#
# URL base for relative URLs pointing at assets
#
my $base_url = 'https://hackerpublicradio.org/';

#
# Text colours
#
my $red     = color('red');
my $green   = color('green');
my $yellow  = color('yellow');
my $magenta = color('magenta');
my $bold    = color('bold');
my $reset   = color('reset');

my %snatts = (
    html_percent => 0,
    html_tags    => 0,
    is_markup    => 0,
);

#
# A list of signature expressions used to try and guess what markup format was
# used.
#
# {{{
#
# Note that the regexes are applied to a string which is a concatenation of
# the notes, so care needs to be taken with anchors. After processing the JSON
# the lines of the notes will be separated by newline characters. Use an 'm'
# qualifier to make a '^' anchor treat it as multi-line.
#
my @markup_types = (
    # Has [...](...) and ![...](...) entities
    {   regex   => qr{!?\[[^\]]+\]\(https?:[^)]+\)},
        type    => ['markdown'],
        comment => "Markdown link or image"
    },

    # Has [...][nnn] entities
    {   regex   => qr{\[[^\]]+\]\[[^\]]+\]},
        type    => ['markdown'],
        comment => "Markdown reference"
    },

    # Has lines beginning with 1-6 hashes (Atx-style header)
    {   regex   => qr{^#{1,6} \w}m,
        type    => ['markdown'],
        comment => "Markdown Atx-style header"
    },

    # A line of stuff followed by a line of = or - (Setex-style header)
    {   regex   => qr{^(\s|\w)+.^[=-]+$}sm,
        type    => ['markdown'],
        comment => "Markdown Setex-style header"
    },

    # Has lines starting with 0+ spaces hyphen 1+ spaces
    {   regex   => qr{^\s*[-*]\s+\w}m,
        type    => [ 'markdown', 'reStructuredText', 'org-mode' ],
        comment => "Markdown or reStructuredText list element"
    },

    # Has lines beginning with '\d.\s' (number(s) dot space(s))
    {   regex   => qr{^\s*\d+\.\s+\w}m,
        type    => [ 'markdown', 'reStructuredText' ],
        comment =>
            "Markdown, reStructuredText or org-mode ordered list element"
    },

    # Has lines which begin with 3+ ` or ~
    {   regex   => qr{^[`~]{3,}$}m,
        type    => ['markdown'],
        comment => "Markdown fenced block delimiter"
    },

    # Has words enclosed in 1 or 2 *
    {   regex   => qr{\*{1,2}\w+\*{1,2}},
        type    => [ 'markdown', 'reStructuredText' ],
        comment => "Markdown or reStructuredText emphasis or bold"
    },

    # Has words enclosed in `
    {   regex   => qr{`\w+`},
        type    => ['markdown'],
        comment => "Markdown verbatim text"
    },

    # Has a line of 3 or more *, - or _ possibly with intervening spaces
    {   regex   => qr{^[*_-]\s?[*_-]\s?[*_-][*_ -]*$}m,
        type    => ['markdown'],
        comment => "Markdown horizontal rule"
    },

    # Has bare http/https URLs
    {   regex   => qr{[^(]\bhttps?:},
        type    => ['txt2tags'],
        comment => "Bare http(s) url"
    },

    # Has [...]_ entities (footnote, citation reference)
    {   regex   => qr{\[(\w+|\d+|#|\*)\]_},
        type    => ['reStructuredText'],
        comment => "reStructuredText footnote, citation or reference"
    },

    # Has `Text <URL>`_ entities (hyperlinks)
    {   regex   => qr{`\w+<https?:[^>]+>`_},
        type    => ['reStructuredText'],
        comment => "reStructuredText hyperlink"
    },

    # Has superscript or subscript markup
    {   regex   => qr{(\^\w+\^|\~\w+\~)},
        type    => ['markdown'],
        comment => "Markdown superscript or subscript"
    },

    # STARS KEYWORD PRIORITY TITLE TAGS is a header definition in Org mode
    {   regex =>
            qr{^\*+\s(TODO|FEEDBACK|VERIFY|\||DONE|DELEGATED)?\s?(\[[#]\w\])?\s?\w+}m,
        type    => ['org-mode'],
        comment => "Org mode header"
    },

    #    {   regex   => qr{},
    #        type    => ['markdown'],
    #        comment => ""
    #    },

);

# }}}

my @markup_found;
my $markup_choice;

my ( $showdir,     $assetdir, $statusfile );
my ( %media_files, $media_files );
my ( %assets,      @assets,    @audio, $sanitised );
my ( $has_audio,   $has_extra, $has_archives );
my ( @archives,    @extracted );
my ( $astate,      $has_pictures, @pictures, $pictures );
my ( $has_markup,  @markup,       $markup );
my ( $json_change, @json_changes, %spellchecks );
my @pstates = (
    'No pictures found',                # 0
    'Pictures that need management',    # 1
    'Managed pictures',                 # 2
);

#
# Enable Unicode mode
#
binmode STDOUT, ":encoding(UTF-8)";
binmode STDERR, ":encoding(UTF-8)";

#-------------------------------------------------------------------------------
# Options and arguments
#-------------------------------------------------------------------------------
my $DEF_DEBUG = 0;

#
# Process options
#
my %options;
Options( \%options );

#
# Default help
#
pod2usage(
    -msg     => "$PROG version $VERSION\n",
    -verbose => 0,
    -exitval => 1
) if ( $options{'help'} );

#
# Detailed help
#
pod2usage(
    -msg       => "$PROG version $VERSION\n",
    -verbose   => 2,
    -exitval   => 1,
    -noperldoc => 0
) if ( $options{'documentation'} );

#
# Collect options
#
my $DEBUG  = ( defined( $options{debug} )  ? $options{debug}  : $DEF_DEBUG );
my $silent = ( defined( $options{silent} ) ? $options{silent} : 0 );
my $test   = ( defined( $options{test} )   ? $options{test}   : 0 );

$showno = $options{episode};
pod2usage(
    -msg     => "$PROG version $VERSION\nMissing mandatory option -episode\n",
    -verbose => 0,
    -exitval => 1
) unless $showno;

my $infile = $options{infile};
pod2usage(
    -msg     => "$PROG version $VERSION\nMissing mandatory option -infile\n",
    -verbose => 0,
    -exitval => 1
) unless $infile;
die "Unable to access input file $infile: $!\n" unless -r $infile;
die "Input file $infile is empty\n" if -z $infile;

#
# Where to put various files. The second group of options must all have
# filenames to be created, but the first can define its own name if not given.
#
my $shownotes = $options{shownotes};

my $formatfile  = $options{format};
my $releasefile = $options{release};
my $assetfile   = $options{assets};
my $picturefile = $options{pictures};
my $zipfile     = $options{zip};

#
# Work on the show notes output file, allowing defaults and substitution points
# for convenience.
#
if ( defined($shownotes) ) {
    $shownotes = output_file_name( $shownotes, $showno, 'hpr%d.out' );
}

#
# Directories and file specific to this show
#
$showdir  = "$basedir/shownotes/hpr$showno";
$assetdir = "$showdir/uploads";
$statusfile = "$showdir/.status";

#-------------------------------------------------------------------------------
# Set up logging keeping the default log layout except for the date. The format
# is "%T [%L] %m" where '%T' is the timestamp, '%L' is the log level and '%m is
# the message.
#-------------------------------------------------------------------------------
my $log = Log::Handler->new();

$log->add(
    file => {
        timeformat     => "%Y/%m/%d %H:%M:%S",
        message_layout => ($test ? 'T ' : '') . "%T [%L] %m",
        filename       => $logfile,
        minlevel       => 0,                     # emergency, emerg
        maxlevel       => 7,                     # debug
        }
);

#-------------------------------------------------------------------------------
#  Read the input file and parse it into $content (hashref)
#-------------------------------------------------------------------------------
my $json = JSON->new->utf8;
open( my $fh, '<:encoding(UTF-8)', $infile );
my $json_text = <$fh>;
close($fh);

my $content;
try {
    $content = decode_json($json_text);
}
catch ($e) {
    warn colored( "Failed to decode the JSON in $infile", 'red' ) . "\n";
    die "Error was: $e\n";
}

$log->info( $showno, "[$VERSION] Processing $infile" );

_debug( $DEBUG > 2, Dumper($content) );

#-------------------------------------------------------------------------------
# Validate the JSON structure in case it's changed for some reason
#-------------------------------------------------------------------------------
die "The JSON doesn't have the right structure\n"
    unless ( exists( $content->{episode} )
    && exists( $content->{metadata} )
    && exists( $content->{host} ) );

print STDERR '-' x 80, "\n";

#-------------------------------------------------------------------------------
# Report details from the parsed data. The percentages of HTML in the Host
# Profile and the notes themselves are computed using an HTML parser that counts
# the tags.
#-------------------------------------------------------------------------------
printf STDERR $ofmt, "Show:", $content->{metadata}{Episode_Number};
printf STDERR $ofmt, "Date:", $content->{metadata}{Episode_Date};

#
# Trim off leading and trailing spaces in these fields
#
$json_change = 0;
for my $key ( 'Title', 'Summary', 'Tags' ) {
    my $str = trim($content->{episode}{$key});
    if ($str ne $content->{episode}{$key}) {
        $content->{episode}{$key} = $str;
        $json_change = 1;
        push(@json_changes,$key);
    }
}

#
# Detect Unicode in the Title, Summary or Tags and flag their presence.
#
for my $key ( 'Title', 'Summary', 'Tags' ) {
    my $uflag = ( $content->{episode}{$key} =~ /[^\x{00}-\x{7F}]/ );
    printf STDERR $ofmt, "$key:",
        colour_if(
            $uflag,
#            'bold yellow on_magenta',
            'black on_bright_yellow',
            $content->{episode}{$key}
        );
}

alert( $ofmt, "JSON to be updated; changes to: " .
    join(',',@json_changes) ) if $json_change;

#
# Check summary length. The field might be filled and something might have
# been lost.
# TODO: Check on other lengths?
#
$summarylength = length($content->{episode}{Summary});
if ($summarylength eq 100) {
    printf STDERR $ofmt, "- Summary: check length";
}

#
# Perform a spelling check on these fields. Collect any that are detected
# then, if any were found, report them.
#
for my $key ( 'Title', 'Summary', 'Tags' ) {
    my @errors = spellcheck( $content->{episode}{$key} );
    $spellchecks{$key} = \@errors if @errors;
}
_debug( $DEBUG > 2, '%spellchecks: ' . Dumper( \%spellchecks ) );
if (%spellchecks) {
    print STDERR colored( "Spelling checks:\n", 'bold yellow' );
    for my $key ( 'Title', 'Summary', 'Tags' ) {
        printf STDERR $ofmt, "- $key:", join( ',', @{ $spellchecks{$key} } )
            if ( exists( $spellchecks{$key} ) );
    }
}

#
# Report on HTML in profile and notes
#
printf STDERR $ofmt, "Profile:",
    count_tags( $content->{host}{Host_Profile} )
    . " HTML start/end tags found";
$snlevel = count_tags( $content->{episode}{Show_Notes} );
printf STDERR $ofmt, "Notes:", $snlevel . " HTML start/end tags found";

#
# Check the note length because if they are too long we might want to split
# them
#
$notelength = length( $content->{episode}{Show_Notes} );
if ( $notelength > $MAXNOTELEN ) {
    printf STDERR $ofmt, "Notes:",
        colored( "Notes are longer than $MAXNOTELEN bytes ($notelength)",
        'bold yellow on_magenta' );
}
elsif ( $notelength <= $MINNOTELEN ) {
    printf STDERR $ofmt, "Notes:",
        colored( "Notes are shorter than $MINNOTELEN bytes ($notelength)",
        'bold yellow on_magenta' );
}

#
# Look for Unicode in the notes and flag it if found
#
if ( find_Unicode( $content->{episode}{Show_Notes} ) ) {
    printf STDERR $ofmt, "Notes:",
        colored( "Unicode characters found", 'bold yellow on_magenta' );
}

#-------------------------------------------------------------------------------
# Always check that the given show number matches what we parsed
#-------------------------------------------------------------------------------
die "The episode specified ($showno) doesn't match the one in the JSON "
    . "($content->{metadata}{Episode_Number})\n"
    unless ( $content->{metadata}{Episode_Number} == $showno );

#
# A zero host id means this is a new host. HOWEVER, it can also mean an
# existing host gave a mail address we don't have in the database. to take
# care!
#
if ( $content->{host}{Host_ID} eq 0 ) {
    print STDERR colored(
        sprintf( $ofmt, "New host:", $content->{host}{Host_Name} ),
        'bold yellow' );
}
else {
    printf STDERR $ofmt, "Host name:", $content->{host}{Host_Name};
}

#
# Save HTML tag count
#
$snatts{html_tags} = $snlevel;

#-------------------------------------------------------------------------------
# Have a go at the HTML tag percentage
#-------------------------------------------------------------------------------
$snatts{html_percent}
    = html_level( $content->{episode}{Show_Notes}, $snlevel );

if ( $snlevel > 0 ) {
    $log->info( $showno, "Notes contain HTML" );
    push( @markup_found, 'HTML' );
    if ( $snatts{html_percent} > 5 ) {
        $log->info( $showno, "Notes probably are HTML" );
        push( @markup_found, 'HTML' );
    }
    else {
        push(
            @markup_found,
            @{  detect_markup(
                    $content->{episode}{Show_Notes}, \@markup_types,
                    $ofmt
                )
            }
        );
    }
}
else {
    $log->info( $showno, "Notes contain no HTML" );
    push(
        @markup_found,
        @{  detect_markup( $content->{episode}{Show_Notes},
                \@markup_types, $ofmt )
        }
    );
}

#-------------------------------------------------------------------------------
# Report what markup was used (not really necessary any more but left in
# because it's interesting).
#-------------------------------------------------------------------------------
# 2022-02-18 Now we have added 'org-mode' which doesn't match any formats sent
# from the web form, so we highlight this to help with the editing process.
# It's possible to use Pandoc to generate Markdown for editing or just convert
# straight to HTML if the result looks good.
#
$markup_choice = choose_markup( \@markup_found );

printf STDERR $ofmt, "Declared format:",
    colour_if(
        $content->{metadata}{Shownotes_Format} ne 'plain_text',
        'bold yellow on_magenta',
        $content->{metadata}{Shownotes_Format}
    );
printf STDERR $ofmt, "Found:", summarise_markup( \@markup_found );
#printf STDERR $ofmt, "Recommendation:",
#   colour_if(
#       $markup_choice eq 'org-mode',
#       'bold yellow on_magenta',
#       $markup_choice
#   );

printf STDERR $ofmt, "Recommendation:",
    colour_switch(
        [
            $markup_choice eq 'org-mode',
            $markup_choice eq 'markdown',
            $markup_choice eq 'txt2tags',
            $markup_choice eq 'none',
            $markup_choice =~ / or /,
        ],
        [
            'bold yellow on_magenta',
            'yellow',
            'yellow',
            'red',
            'yellow',
        ],
        $markup_choice
    );

#-------------------------------------------------------------------------------
# What about the media?
#-------------------------------------------------------------------------------
# Examine the delivered media details. Start by finding the maximum index
# (media_files is a JSON hash containing keys like 'name' where the values
# are arrays). The elements are linked by their positions.
#
my $maxind = $#{ $content->{metadata}{FILES}->{media_files}->{name} };

#
# Step through the media file names and their sizes building a Perl hash keyed by the
# name containing the file size of each
#
for ( my $i = 0; $i <= $maxind; $i++ ) {
    $media_files{ $content->{metadata}{FILES}->{media_files}->{name}->[$i] }
        = $content->{metadata}{FILES}->{media_files}->{size}->[$i];
}

#
# If a media file is zero size, remove it from the list
#
for my $key ( keys(%media_files) ) {
    if ( $media_files{$key} eq 0 ) {
        delete( $media_files{$key} );
        $maxind--;
    }
}

#
# Generate a string containing the names of the (non-empty) media files
#
if (%media_files) {
    $media_files = join( ", ", keys(%media_files) );
}

#-------------------------------------------------------------------------------
# Determine the 'asset' state of the show
#-------------------------------------------------------------------------------
#
# (The information is in %media_files although strictly there could be other
# files like scripts, and pictures.) The audio is in %media_files unless the
# host has sent in an URL for downloading.
#
if (%media_files) {
    #
    # Build a hash (%assets) keyed by filename with the value being a hashref
    # with keys 'type' and 'size'.
    #
    for ( my $i = 0; $i <= $maxind; $i++ ) {
        my $name = $content->{metadata}{FILES}->{media_files}->{name}->[$i];
        $assets{$name} = {
            type => $content->{metadata}{FILES}->{media_files}->{type}->[$i],
            size => $content->{metadata}{FILES}->{media_files}->{size}->[$i],
        };
    }

    #
    # Process the assets
    #
    if (%assets) {
        #
        # Save just the file names in an array (@assets) for convenience.
        #
        @assets = keys(%assets);

        #-------------------------------------------------------------------------------
        # Check that there's audio in the media.
        #-------------------------------------------------------------------------------
        #
        # First look for any assets marked as audio or video (this one because
        # some audio has been seen to be marked this way). We haven't expanded
        # archive files yet because we don't want to eliminate those files
        # from @assets.
        #
        $has_audio = @audio
            = grep { $assets{$_}->{type} =~ /^(audio|video)/i } keys(%assets);

        #
        # Now, if no audio has yet been found, look for audio file extensions
        #
        unless ($has_audio) {
            $has_audio = @audio
                = grep { $_ =~ /\.(aac|flac|mp3|ogg|spx|wav)$/i }
                keys(%assets);
        }

        #
        # Check there is not more than one audio file
        #
        if ($#audio > 1) {
            print colored(
                "**Warning** More than one audio file uploaded", 'red'), "\n";
            $log->warning( $showno, "More than one audio file uploaded" );
        }

        #
        # We have audio. Use the @audio array to remove the matches from
        # @assets.
        #
        if ($has_audio) {
            @assets = array_difference( \@assets, \@audio );
        }
        if (@assets) {
            _debug( $DEBUG > 2, '@assets: ' . Dumper( \@assets ) );
        }

        #-------------------------------------------------------------------------------
        # Look for un-processed markup in the assets hash. For the moment we
        # only look for Markdown.
        #-------------------------------------------------------------------------------
        $has_markup = @markup
            = grep { $assets{$_}->{type} =~ m{^text/markdown$} } keys(%assets);
        $markup = join( ', ', @markup );
        if (@markup) {
            _debug( $DEBUG > 2, '@markup ' . Dumper( \@markup ) );
        }

        #-------------------------------------------------------------------------------
        # Look for archive files in the assets
        #-------------------------------------------------------------------------------
        if ( find_archive( \%assets, \@archives ) > 0 ) {
            _debug( $DEBUG > 1, '@archives: ' . Dumper( \@archives ) );

            #
            # Extract the contents of any archives into the show directory and
            # write details into %assets and @extracted.
            #
            $has_archives
                = ( extract_archives( \%assets, \@archives, \@extracted, $showdir ) > 0 );
        }
        else {
            #
            # No archives, but we may have assets. Keep copies of the assets
            # in case we need to make destructive changes.
            #
            if (@assets && $zipfile) {
                my @input = map {"$showdir/$_"} @assets;
                _debug( $DEBUG > 2, 'Zip creation - @input: ' . Dumper(
                        \@input ) );

                zip \@input => $zipfile
                    or die "Zip creation failed: $ZipError\n";
            }
        }

        #-------------------------------------------------------------------------------
        # Refresh assets after expanding archives
        #-------------------------------------------------------------------------------
        #
        # Remove archive files and add extracted files to the assets. We may
        # have a different number than we started with because any archives found
        # will have been expanded and the archive names removed.
        #
        @assets = array_difference(\@assets,\@archives);
        push(@assets, @extracted);

        #
        # Remove markup files from the assets so we don't upload them
        #
        @assets = array_difference(\@assets,\@markup);

        #
        # Remove directory stuff from @assets elements and %assets keys
        #
        # foreach my $file (@assets) {
        #     ( my $key = $file ) =~ s{^.*/}{};
        #     $assets{$key} = delete( $assets{$file} );
        #     $file = $key;
        # }
        prepare_assets( \%assets, \@assets, $showdir, $assetdir );
        if (@assets) {
            _debug( $DEBUG > 1, 'After prepare_assets, %assets: ' . Dumper( \%assets ) );
            _debug( $DEBUG > 1, 'After prepare_assets, @assets: ' . Dumper( \@assets ) );
        }

        #-------------------------------------------------------------------------------
        # Sanitise any filenames that need it
        #-------------------------------------------------------------------------------
        $sanitised = sanitise_filenames( \%assets, \@assets, $assetdir );

        #-------------------------------------------------------------------------------
        # Look for any image files in the assets hash
        #-------------------------------------------------------------------------------
        $has_pictures = @pictures
            = grep { $assets{$_}->{type} =~ /^image/ } keys(%assets);
        $pictures = join( ', ', @pictures );

        # if (@archives =
        #     grep { $assets{$_}->{type} =~ q{^application/zip} } keys(%assets)) {
        #     my $archive = Archive::Any->new( "$showdir/$archives[0]" );
        #     @archived = $archive->files;
        #     $has_pictures += scalar(@archived);
        # }

        # print "D> Pictures found\n" if $has_pictures;
        # if ($has_pictures && @archived && $DEBUG > 2) {
        #    if ($has_pictures && $DEBUG > 2) {
        #        _debug (1, '@archives: ' . Dumper(\@archives));
        #        _debug (1, '@archived: ' . Dumper(\@archived));
        #    }

        if ( $has_pictures && $DEBUG > 2 ) {
            _debug( 1, '@pictures: ' . Dumper( \@pictures ) );
        }

        #
        # Look for files not matching audio or image
        #
        $has_extra = grep { $assets{$_}->{type} !~ /^(image|audio|video)/ }
            keys(%assets);

        _debug( $DEBUG > 2, '%assets: ' . Dumper( \%assets ) ) if (%assets);

        $log->info( $showno, "Media files: $media_files" ) if ($media_files);
        $log->info( $showno, "Pictures: " . join( ', ', @pictures ) )
            if (@pictures);
        $log->info( $showno, "Markup " . join( ', ', @markup ) )
            if (@markup);
        $log->info( $showno, "Assets: " . join( ', ', @assets ) )
            if (@assets);

    }
}
else {
    #
    # No media files at all!
    #
    $has_archives = $has_pictures = $has_audio = $has_extra = 0;
}

#-------------------------------------------------------------------------------
# Sanity checks
#-------------------------------------------------------------------------------
#
# 1. Check for included media or a download URL. Alert if neither
#
if ($media_files) {
    #
    # Have media, but do we have audio?
    #
    # printf STDERR $ofmt, "Media files:", $media_files;
    printf STDERR "%s\n",
        textFormat( $media_files, 'Media files:', 'L', 18, 80 );
    unless ($has_audio) {
        alert( $ofmt, "No audio!" );
        $log->error( $showno, "No audio!" );
    }
}
elsif ( length( $content->{metadata}{url} ) > 0 ) {
    #
    # No media, but do we have a download URL?
    #
    printf STDERR $ofmt, "Download from:", $content->{metadata}{url};
}
else {
    #
    # No media and no download URL, so no audio
    #
    alert( $ofmt, "No audio!" );
    $log->error( $showno, "No audio!" );
}

#
# 2. Alert if no tags
#
if ( length( $content->{episode}{Tags} ) == 0 ) {
    alert( $ofmt, "No tags!" );
    $log->error( $showno, "No tags!" );
}

#
# 3. Declared html5 but no HTML found
#
if ( $snlevel == 0 && $content->{metadata}{Shownotes_Format} =~ /html5/i ) {
    alert( $ofmt, "Declared format 'html5' with no HTML in the notes!" );
    $log->error( $showno,
        "Declared format 'html5' with no HTML in the notes!" );
    #    printf STDERR $ofmt, "",
    #        "${yellow}Forced format to 'plain_text'.${reset}";
    printf STDERR colored(
        sprintf( $ofmt, "", "Forced format to 'plain_text'." ),
        'bold yellow' );
    $content->{metadata}{Shownotes_Format} = 'plain_text';
    # TODO: propagate the format change into the JSON
}

#
# 4. Tags aren't CSV. Not much of a check since so many formats are acceptable
#    here but not in the database.
#    If there's only one tag write an alert reporting the length in case the
#    sender forgot the commas.
#    TODO: Reconsider after using for a while. May be no point in doing this.
#
if ( length( $content->{episode}{Tags} ) > 0 ) {
    my $csv = Text::CSV->new( { sep_char => ',', allow_whitespace => 1 } )
        or warn "Problem with Text::CSV: $@";
    my $status = $csv->parse( $content->{episode}{'Tags'} );
    unless ($status) {
        alert( $ofmt, $csv->error_input() . "\n" . $csv->error_diag() );
        $log->error( $showno, "Tag format is invalid" );
    }
    my @cols = $csv->fields();
    if ( scalar(@cols) eq 1 ) {
        alert( $ofmt,
            "Only one tag " . length( $cols[0] ) . " characters long!" );
    }
}

#
# 5. Declared HTML, but does it have assets and if so are they properly
#    pointed to?
#
if ( $snlevel > 0 && $content->{metadata}{Shownotes_Format} =~ /html5/i ) {
    #
    # Perform a check on the notes in memory, looking for asset links and
    # checking they are kosher
    #
    unless ( check_html( $content->{episode}{Show_Notes}, $base_url ) ) {
        print colored(
            "Apparently incorrect URLs detected in the notes", 'red'
            ), "\n";
    }
}

#
# 6. Declared plain text, but analysis recommends HTML, so highlight this to
#    allow changes to be made.
#
#if ( $snlevel > 5 && $content->{metadata}{Shownotes_Format} =~ /plain_text/i ) {
if (   $markup_choice eq 'html5'
    && $content->{metadata}{Shownotes_Format} =~ /plain_text/i )
{
    alert( $ofmt,
        "Declared format 'plain_text' but notes seem to be HTML5!" );
    $log->error( $showno,
        "Declared format 'plain_text' but notes seem to be HTML5!" );
}

#
# 7. The host has sent in markup version(s) of their external notes, so we
#    need to take special action.
#
if ($markup) {
    printf STDERR "%s\n",
        textFormat( $markup, 'Markup files:', 'L', 18, 80 );
}

#-------------------------------------------------------------------------------
# Determine the picture asset state
#-------------------------------------------------------------------------------
#
# If there are pictures and the format is plain text then the state is 'Pictures
# that need management' (1), otherwise 'Managed pictures' (2). If no pictures
# then state is 'No pictures found' (0).
#
if ($has_pictures) {
    $astate
        = ( $content->{metadata}{Shownotes_Format} eq 'plain_text' ? 1 : 2 );
    # printf STDERR $ofmt, "Pictures:", $pictures;
    printf STDERR "%s\n", textFormat( $pictures, 'Pictures:', 'L', 18, 80 );
}
else {
    $astate = 0;
}
printf STDERR $ofmt, "Picture state:",
    colour_if( $astate > 0, 'bold yellow on_magenta', $pstates[$astate] );

#-------------------------------------------------------------------------------
# Flag other assets so they aren't missed
#-------------------------------------------------------------------------------
if ($has_extra) {
    printf STDERR $ofmt, "Asset state:",
        colored( ['bold yellow on_magenta'], 'Other files uploaded' );
}

if (@assets) {
    printf STDERR $ofmt, "Archives: ", join( ', ', @archives )
        if $has_archives;
    # printf STDERR $ofmt, "All assets:", join( ', ', @assets );
    printf STDERR "%s\n",
        textFormat( join( ', ', @assets ), 'All assets:', 'L', 18, 80 );
    if ( $sanitised > 0 ) {
        printf STDERR $ofmt, 'Assets sanitised:', "$sanitised";
    }
}

#
# End of content report
#
print STDERR '-' x 80, "\n";

#
# Log choices and problems
#
$log->info( $showno,
    "Declared format: " . $content->{metadata}{Shownotes_Format} );
$log->info( $showno, "Recommended markup processing: $markup_choice" );


#-------------------------------------------------------------------------------
# Output to relevant files
#-------------------------------------------------------------------------------
#
# Write the declared format to a file if requested
#
if ( defined($formatfile) ) {
    open( my $fmtfile, '>:encoding(UTF-8)', $formatfile )
        or die "Unable to open output file $formatfile $!\n";
    print $fmtfile $content->{metadata}{Shownotes_Format}, "\n";
    close($fmtfile);
    print colored( "Format details written to $formatfile", 'green' ), "\n"
        unless $silent;
    $log->info( $showno, "Format details written to $formatfile" );
}

#
# Write the release date to a file if requested
#
if ( defined($releasefile) ) {
    open( my $relfile, '>:encoding(UTF-8)', $releasefile )
        or die "Unable to open output file $releasefile $!\n";
    print $relfile $content->{metadata}{Episode_Date}, "\n";
    close($relfile);
    print colored( "Release date written to $releasefile", 'green' ), "\n"
        unless $silent;
    $log->info( $showno, "Release date written to $releasefile" );
}

#
# Write the picture file names to a file if requested and there are some
#
if ( @pictures && defined($picturefile) ) {
    open( my $picfile, '>:encoding(UTF-8)', $picturefile )
        or die "Unable to open output file $picturefile $!\n";

    #
    # Sort the pictures alphabetically in case the host didn't order them
    # sensibly.
    #
    foreach my $picture (sort(@pictures)) {
        print $picfile "$picture\n";
    }
    close($picfile);

    print colored( "Picture names written to $picturefile", 'green' ), "\n"
        unless $silent;
    $log->info( $showno, "Picture names written to $picturefile" );
}

#
# Write the asset file names to a file if requested and there are some
#
if ( @assets && defined($assetfile) ) {
    open( my $assfile, '>:encoding(UTF-8)', $assetfile )
        or die "Unable to open output file $assetfile $!\n";
    foreach my $asset (@assets) {
        print $assfile "$asset\n";
    }
    close($assfile);
    print colored( "Asset names written to $assetfile", 'green' ), "\n"
        unless $silent;
    $log->info( $showno, "Asset names written to $assetfile" );
}

#-------------------------------------------------------------------------------
# Write the shownotes component to a file if requested
#-------------------------------------------------------------------------------
if ($shownotes) {
    #
    # Remove carriage returns
    #
    $content->{episode}{Show_Notes} =~ s/\r//g;
    _debug( $DEBUG > 2, quote_crlf( $content->{episode}{Show_Notes} ) );

    #
    # Save the notes
    #
    open( my $out, '>:encoding(UTF-8)', $shownotes )
        or die "Unable to open output file $shownotes: $!\n";
    print $out $content->{episode}{Show_Notes}, "\n";
    close($out);
    print colored( "Shownotes written to $shownotes", 'green' ), "\n"
        unless $silent;
    $log->info( $showno, "Shownotes written to $shownotes" );

    #
    # We're generating shownotes and we have declared HTML (or we're pretty
    # sure after parsing that they are HTML), so save the notes as HTML as
    # well. Now achieved simply by symlinking the two files if they are
    # different. This allows the 'do_edit` script to edit the *.out file and
    # thereby affect the *.html file.
    #
    # 2018-09-13: Removed the check for $markup_choice because it's adding
    # links inappropriately.
    #
    # 2020-05-31: Turned the symlink to a hard link
    #
    if ( $content->{metadata}{Shownotes_Format} =~ /html5/i ) {
        ( my $htmlfile = $shownotes ) =~ s/\.[^.]+$/.html/;

        unless ( $htmlfile eq $shownotes ) {
            if ( link( $shownotes, $htmlfile ) ) {
                print colored(
                    "HTML notes - linked $htmlfile to $shownotes", 'green'
                    ),
                    "\n";
                $log->info( $showno,
                    "HTML notes - linked $htmlfile to $shownotes" );
            }
            else {
                print colored(
                    "**Warning** Failed to link $htmlfile to $shownotes",
                    'red'
                    ),
                    "\n";
                print colored( "Equivalent CLI command:", 'yellow' ),
                    "\n",
                    colored( "ln $shownotes $htmlfile", 'yellow' ), "\n";
                $log->warning( $showno,
                    "Failed to link $htmlfile to $shownotes" );
            }
        }
    }
}
else {
    print colored( "No shownotes written", 'green' ), "\n" unless $silent;
    $log->info( $showno, "No shownotes written" );
}

#
# Save the status
#
update_status($statusfile,'parsed');

exit;

#===  FUNCTION  ================================================================
#         NAME: find_archive
#      PURPOSE: Look though the assets for files that are some kind of archive
#   PARAMETERS: $assets         hashref pointing to the asset filename list
#                               (contains size and type information)
#               $archfiles      arrayref to hold the archive files found
#      RETURNS: Number of archive files found
#  DESCRIPTION:
#       THROWS: No exceptions
#     COMMENTS: Uses 'any' from 'List::MoreUtils'
#     SEE ALSO: N/A
#===============================================================================
sub find_archive {
    my ( $assets, $archfiles ) = @_;

    my ( $count, $type );

    #
    # Return if no assets given
    #
    return 0 unless %$assets;

    #
    # Known archive types (from what's been seen so far). Declared as
    # a 'state' variable so it's only initialised once.
    #
    state @mime_types = (
        #{{{
        'application/x-tar',
        'application/x-bzip-compressed-tar',
        'application/x-compressed-tar',
        'application/x-gtar',
        'application/zip',
        'application/x-zip-compressed',
        'application/x-bzip2',
        'application/gzip',
        #}}}
    );

    $count = 0;
    foreach my $file ( keys(%$assets) ) {
        $type = $assets->{$file}->{type};
        if ( any { $_ eq $type } @mime_types ) {
            push( @$archfiles, $file );
            $count++;
        }
    }

    return $count;
}

#===  FUNCTION  ================================================================
#         NAME: extract_archives
#      PURPOSE: Given a list of archive files, extract their contents and save
#               the file names
#   PARAMETERS: $assets         hashref containing the filename list and some
#                               attributes. Read and written.
#               $archfiles      arrayref holding the archive files found (by
#                               find_archive). Read only.
#               $extracted      arrayref to hold the files extracted from the
#                               archives.
#               $showdir        path to find the archive file(s)
#      RETURNS: Number of files extracted. Alters %$assets to reflect the new
#               files and remove the extracted archive files.
#  DESCRIPTION: Just return with zero if there are no archives.
#               Process the archives one by one. Open the current one with
#               Archive::Any and save the file contents to return to the
#               caller. Extract the archive contents to the 'uploads'
#               directory in the directory for the show.
#               Remove all the archive file references from %$assets. Add in
#               the extracted files using MIME::Types to get the mime details
#               and 'stat' to get the sizes. Keep a cound of the extracted
#               files to return to the caller.
#       THROWS: No exceptions
#     COMMENTS: None
#     SEE ALSO: N/A
#===============================================================================
sub extract_archives {
    my ( $assets, $archfiles, $extracted, $showdir ) = @_;

    my ( $aa, $mt, $typename, $filecount );

    #
    # Return if no archives given
    #
    return 0 unless @$archfiles;

    #
    # Loop through the archives we were given. There could be more than one,
    # though it's likely to be rare.
    #
    foreach my $arch (@$archfiles) {
        #print "D> Archive file: $arch\n";

        #
        # Open the archive file
        #
        $aa = Archive::Any->new("$showdir/$arch");

        #
        # Get and store a list of the contents of the archive
        #
        push( @$extracted, $aa->files );

        #
        # Extract everything to the show directory
        #
        #print "About to extract to $assetdir\n";
        $aa->extract($showdir);
    }

    #print "D> Extracted files: ",join("\nD> ", @$extracted),"\n";

    #
    # Now update the %$assets hash, first deleting the archive files we have
    # extracted the contents of.
    #
    foreach my $arch (@$archfiles) {
        delete( $assets->{$arch} );
    }
    # _debug(1,'%$assets after deletions: ' . Dumper($assets));

    #
    # Now add in the extracted files with their details
    #
    foreach my $asset (@$extracted) {
        #
        # Rationalise the filename
        #
        $asset = canonpath($asset);

        #
        # Get the MIME type of the file (stringifying the result of mimeTypeOf
        # otherwise it's a MIME::Type object)
        #
        $mt               = MIME::Types->new;
        $typename         = "" . $mt->mimeTypeOf("$showdir/$asset");
        $assets->{$asset} = {
            type => $typename,
            size => ( stat("$showdir/$asset") )[7],
        };

        $filecount++;
    }
    # _debug(1,'%$assets after additions: ' . Dumper($assets));

    return $filecount;
}

#===  FUNCTION  ================================================================
#         NAME: array_difference
#      PURPOSE: Determine the members of the left array not in the right array
#   PARAMETERS: $left           arrayref to the left-hand array
#               $right          arrayref to the right-hand array
#      RETURNS: The difference between the two arrays (@$left - @$right in
#               concept). All the elements that are in @$left but not in
#               @$right are returned.
#  DESCRIPTION: Loads a hash with keys from the $left array. This will remove
#               duplicates as a by-product. Using the $right array to provide
#               keys we delete all hash elements that match. This leaves the
#               hash containing the difference between the two arrays by
#               "subtracting" the right from the left. Another consequence of
#               the method used is that the resulting list is not in the same
#               order as the $left array.
#       THROWS: No exceptions
#     COMMENTS: None
#     SEE ALSO: N/A
#===============================================================================
sub array_difference {
    my ( $left, $right ) = @_;

    my %files;

    unless (ref($left) eq 'ARRAY' && ref($right) eq 'ARRAY') {
        warn "Arguments to array_difference must both be arrayrefs\n";
        return;
    }

    @files{@$left} = ();          # All files are the keys.
    delete( @files{@$right} );    # Remove the links.

    return keys(%files);
}

#===  FUNCTION  ================================================================
#         NAME: prepare_assets
#      PURPOSE: Tidy assets, removing directories if necessary, and place the
#               files in the upload directory
#   PARAMETERS: $assets_h       hashref containing asset details
#               $assets_a       arrayref containing asset names
#               $showdir        string containing the show directory
#               $assetdir       string containing the asset upload directory
#                               path
#      RETURNS: Nothing
#  DESCRIPTION: 
#       THROWS: No exceptions
#     COMMENTS: None
#     SEE ALSO: N/A
#===============================================================================
sub prepare_assets {
    my ( $assets_h, $assets_a, $showdir, $assetdir ) = @_;

    my ( $file, $dir, $base, @dels, %dirs, %moves );

    #
    # Create the asset directory
    #
    unless ( -e $assetdir ) {
        mkdir($assetdir) or die "Failed to create $assetdir\n";
    }

    #
    # Remove directory stuff from @assets elements and %assets keys saving
    # directories, and moving files as we go.
    #
    # foreach my $file (@$assets_a) {
    for ( my $i = 0; $i <= $#{$assets_a}; $i++ ) {
        #
        # Get and parse an asset filename (using Path::Class). Force the
        # directory and base part to become strings ("stringify")
        #
        $file = file( $assets_a->[$i] );
        $dir  = "" . $file->dir;
        $base = "" . $file->basename;

        #
        # Does the asset have a directory part and a basename part? If so we
        # want to move the file to the asset directory and eventually delete
        # the directory the asset came from.
        #
        # if ( ( $dir, $base ) = ( $file =~ qr{^(.*/)(.+)$} ) ) {
        if ( $dir !~ /^\.?$/ && $base ne '' ) {
            #
            # The asset filename starts with a directory.
            # Save it.
            #
            $dirs{$dir} = 1 unless exists( $dirs{$dir} );
            # push( @dirs, $dir );

            #
            # Update the hash key
            #
            $assets_h->{$base} = delete( $assets_h->{$assets_a->[$i]} );

            #
            # Save the move
            #
            $moves{"$showdir/$file"} = "$assetdir/$base";

            #
            # Alter the asset name in @assets
            #
            # $file = $base;
            $assets_a->[$i] = $base;
        }
        elsif ( $base eq '' ) {
            #
            # We just have a directory, no basename, so queue to delete it
            #
            push(@dels,$i);
        }
        else {
            #
            # No directory (or just '.'), just a basename, so queue the move
            #
            $moves{"$showdir/$base"} = "$assetdir/$base";
        }

    }

    #
    # Perform any moves (using File::Copy)
    #
    foreach my $move (keys(%moves)) {
        # TODO: Is the 'unless' needed?
        move ($move,$moves{$move}) unless ($move =~ qr{/$});
    }

    #
    # Clean out unwanted directories from @$assets_a
    #
    foreach my $del (@dels) {
        splice(@$assets_a,$del,1);
    }

    #
    #
    # Remove any real directories created by unpacking archives
    #
    foreach my $d (keys(%dirs)) {
        rmdir("$showdir/$d") or
            warn "Problem deleting unwanted directory: $d\n";
    }

}

#===  FUNCTION  ================================================================
#         NAME: sanitise_filenames
#      PURPOSE: Deal with filenames with spaces to make life easier
#   PARAMETERS: $assets_h       hashref containing asset details
#               $assets_a       arrayref containing asset names
#               $assetdir       string containing the asset upload directory
#                               path
#      RETURNS: Number of fixes
#  DESCRIPTION: 
#       THROWS: No exceptions
#     COMMENTS: None
#     SEE ALSO: N/A
#===============================================================================
sub sanitise_filenames {
    my ( $assets_h, $assets_a, $assetdir ) = @_;

    my ( $file, $newfile, %moves, $changes );

    #
    # Look at all asset filenames
    #
    for ( my $i = 0; $i <= $#{$assets_a}; $i++ ) {
        $file = $assets_a->[$i];

        #
        # Does the name have whitespace?
        #
        if ($file =~ /\s/) {
            #
            # Sanitise and save the change
            #
            ($newfile = $file) =~ s/ /_/g;

            #
            # Adjust asset array and hash
            #
            $assets_a->[$i] = $newfile;
            $assets_h->{$newfile} = delete( $assets_h->{$file} );

            #
            # Save the file move
            #
            $moves{"$assetdir/$file"} = "$assetdir/$newfile";
        }
    }

    if (%moves) {
        #
        # Perform any moves (using File::Copy)
        #
        $changes = 0;
        foreach my $move (keys(%moves)) {
            move ($move,$moves{$move});
            $changes++;
        }

        return $changes;
    }
    else {
        return 0;
    }

}

#===  FUNCTION  ================================================================
#         NAME: choose_markup
#      PURPOSE: Given a list of possible markup types, choose which one to
#               recommend
#   PARAMETERS: $markup_found   Arrayref containing markup names
#      RETURNS: The markup name to use
#  DESCRIPTION: The array referenced by $markup_found should have one or more
#               instances of a markup type. We want to decide which one to
#               recommend for processing.
#       THROWS: No exceptions
#     COMMENTS: None
#     SEE ALSO: N/A
#===============================================================================
sub choose_markup {
    my ($markup_found) = @_;

    my %freq;
    my @res;

    return 'none' unless $markup_found && scalar(@$markup_found) > 0;
    return 'none' if $markup_found->[0] eq 'none';

    #
    # Collect frequencies of each markup type
    #
    foreach my $mu (@$markup_found) {
        if ( exists( $freq{$mu} ) ) {
            $freq{$mu}++;
        }
        else {
            $freq{$mu} = 1;
        }
    }

    #
    # If there's only one key then that's the recommendation
    #
    return ( keys(%freq) )[0] if scalar( keys(%freq) ) == 1;

    #
    # Find the highest frequency
    #
    my $max = 0;
    foreach my $key ( keys(%freq) ) {
        if ( $freq{$key} > $max ) {
            $max = $freq{$key};
        }
    }

    #
    # Collect all the keys with the maximum frequency
    #
    foreach my $key ( keys(%freq) ) {
        if ( $freq{$key} == $max ) {
            push( @res, $key );
        }
    }

    #
    # Return the matching markup types
    #
    return join( ' or ', @res );

}

#===  FUNCTION  ================================================================
#         NAME: summarise_markup
#      PURPOSE: Summarise the list of matches to markup types
#   PARAMETERS: $markup_found   Arrayref containing markup names
#      RETURNS: A string containing markup types and their frequencies
#  DESCRIPTION:
#       THROWS: No exceptions
#     COMMENTS: None
#     SEE ALSO: N/A
#===============================================================================
sub summarise_markup {
    my ($markup_found) = @_;

    my %freq;
    my @res;

    return 'none' unless $markup_found && scalar(@$markup_found) > 0;
    return 'none' if $markup_found->[0] eq 'none';

    #
    # Collect frequencies of each markup type
    #
    foreach my $mu (@$markup_found) {
        if ( exists( $freq{$mu} ) ) {
            $freq{$mu}++;
        }
        else {
            $freq{$mu} = 1;
        }
    }

    foreach my $key ( sort( keys(%freq) ) ) {
        push( @res, "$key($freq{$key})" );
    }

    return join( ", ", @res );

}

#===  FUNCTION  ================================================================
#         NAME: detect_markup
#      PURPOSE: Tries to detect whether there is any markup in the string and
#               if so what it is
#   PARAMETERS: $string         String to examine
#               $markup_types   Arrayref pointing to regular expression
#                               structures to be used in the detection process
#               $ofmt           Output format for 'printf'
#      RETURNS: A string containing the markup type(s) or 'none'
#  DESCRIPTION: Applies regular expressions to the string (notes) looking for
#               markup signatures. Array @markup_types contains one hash per
#               regex, along with an array of markdown types the regex relates
#               to and a comment explaining what's being detected. The regex
#               is matched against the string and a count of matches returned.
#               If there are matches then an analysis is printed on STDERR.
#               The variable MARKUP_DETECTED is a global variable accessed
#               through 'our' which controls whether the analyses have
#               a single header.
#       THROWS: No exceptions
#     COMMENTS: None
#     SEE ALSO: N/A
#===============================================================================
sub detect_markup {
    my ( $string, $markup_types, $ofmt ) = @_;

    my ( @type, $count );
    our $MARKUP_DETECTED;

    return ['none'] unless $string;

    foreach my $test (@$markup_types) {
        if ( $count = () = ( $string =~ /$test->{regex}/g ) ) {
            print STDERR "Markup check:\n" if ( $MARKUP_DETECTED++ == 0 );
            printf STDERR $ofmt, "- Matched:",
                $test->{regex} . " -> " . join( ",", @{ $test->{type} } );
            printf STDERR $ofmt, "  Comment:", $test->{comment};
            printf STDERR $ofmt, "  Matches:", $count;
            push( @type, @{ $test->{type} } ) for 1 .. $count;
        }
    }

    return ( @type ? \@type : ['none'] );
}

#===  FUNCTION  ================================================================
#         NAME: count_tags
#      PURPOSE: Counts the number of start tags in what might be HTML
#   PARAMETERS: $string         String to examine
#      RETURNS: An integer count of the number of start tags found
#  DESCRIPTION: Uses HTML::Parser to parse the input string. A handler is
#               defined which accumulates start and end tags in an array. The
#               number of start tags found is returned to the caller.
#       THROWS: No exceptions
#     COMMENTS: None
#     SEE ALSO: N/A
#===============================================================================
sub count_tags {
    my ($string) = @_;

    my @accum;
    chomp($string);
    return 0 unless $string;

    my $p = HTML::Parser->new(
        api_version => 3,
        handlers    => {
            start => [ \@accum, "event,text" ],
            end   => [ \@accum, "event,text" ]
        }
    );
    $p->parse($string);
    $p->eof;

    return scalar(@accum);
}

#===  FUNCTION  ================================================================
#         NAME: html_level
#      PURPOSE: Returns the proportion of HTML in a string
#   PARAMETERS: $string         String to examine
#               $tag_count      Number of start & end tags
#      RETURNS: A floating point number representing the (very) rough
#               percentage of HTML tags in the string
#  DESCRIPTION:
#       THROWS: No exceptions
#     COMMENTS: None
#     SEE ALSO: N/A
#===============================================================================
sub html_level {
    my ( $string, $tag_count ) = @_;

    return '0.0' unless $string;

    chomp($string);
    my @words = split( /\s+/, $string );
    my $wc    = scalar(@words);

    return sprintf( "%.2f", ( $tag_count / $wc ) * 100 );
}

#===  FUNCTION  ================================================================
#         NAME: check_html
#      PURPOSE: Checks an HTML string to ensure HPR links are valid
#   PARAMETERS: $html           HTML string
#      RETURNS: True if valid, otherwise false
#  DESCRIPTION:
#       THROWS: No exceptions
#     COMMENTS: None
#     SEE ALSO: N/A
#===============================================================================
sub check_html {
    my ( $html, $base_url ) = @_;

    my ( $parser, @links, $result, $etype, $aname, $avalue );
    my ( $base_re, $show_re, $host_re, $anchor_re, $asset_re );

    chomp($html);
    return 1 unless $html;

    $parser = HTML::LinkExtor->new( undef, $base_url );
    $parser->parse($html);
    @links = $parser->links;

    #    print Dumper( \@links ), "\n";

    #
    # Make some regular expressions. Note that the $base_url is expected to
    # end with a '/'
    #
    $base_re   = qr{^${base_url}};
    $host_re   = qr{${base_re}correspondents.php\?hostid=\d{1,4}$};
    $show_re   = qr{${base_re}eps.php\?id=\d{1,4}$};
    $anchor_re = qr{${base_re}\#.+$};
    $asset_re  = qr{${base_re}eps/hpr\d{1,4}/.+$};

    #
    # Walk the collected links
    #
    $result = 0;
    foreach my $element (@links) {
        #
        # Collect the parsed links
        #
        ( $etype, $aname, $avalue ) = @$element;

        #
        # Only check HPR links
        #
        if ( $avalue =~ /$base_re/ ) {

            #
            # Not interested in correctly formed asset or show links
            #
            return 1
                if ( $avalue =~ /$show_re|$host_re|$anchor_re|$asset_re/ );

            #
            # Something seems to be wrong with a link
            #
            if ( $etype eq 'a' ) {
                $result++;
                print "<A> tag $avalue\n" unless $silent;
            }
            elsif ( $etype eq 'img' ) {
                $result++;
                print "<A> tag $avalue\n" unless $silent;
            }

        }
    }

    return $result == 0;
}

#===  FUNCTION  ================================================================
#         NAME: find_Unicode
#      PURPOSE: Scans a text string for Unicode in case we want to remove it
#   PARAMETERS: $string         String to examine
#      RETURNS: True (1) if Unicode is found, otherwise False (0)
#  DESCRIPTION:
#       THROWS: No exceptions
#     COMMENTS: None
#     SEE ALSO: N/A
#===============================================================================
sub find_Unicode {
    my ($string) = @_;

    return ( $string =~ /[^\x{00}-\x{7F}]/ );
}

#===  FUNCTION  ================================================================
#         NAME: trim
#      PURPOSE: Trims leading and trailing spaces from a string
#   PARAMETERS: string          string to trim
#      RETURNS: Trimmed string
#  DESCRIPTION: 
#       THROWS: No exceptions
#     COMMENTS: None
#     SEE ALSO: N/A
#===============================================================================
sub trim {
    my ($str) = @_;

    $str =~ s/^\s+|\s+$//g;
    return $str;
}

#===  FUNCTION  ================================================================
#         NAME: output_file_name
#      PURPOSE: Generate an output file name with three choices
#   PARAMETERS: $optarg         the argument to the option choosing the filename
#               $showno         the show number to add to certain name types
#               $template       a default 'sprintf' template for the name
#      RETURNS: The name of the output file
#  DESCRIPTION: If there's a defined output filename then there are three
#               options: a null string, a plain filename and a substitution
#               string with '%d' sequences. The null string means the user used
#               '-option' without a value, so we want to generate a substitution
#               string. A string with '%d' requires a check to ensure there's
#               the right number, just one. The plain filename needs no more
#               work.
#       THROWS: No exceptions
#     COMMENTS: None
#     SEE ALSO: N/A
#===============================================================================
sub output_file_name {
    my ( $optarg, $showno, $template ) = @_;

    my ( $filename, $count );

    #
    # We shouldn't be called with a null option argument
    #
    return unless defined($optarg);

    #
    # Does the option have an argument?
    #
    if ( $optarg =~ /^$/ ) {
        #
        # No argument; use the show number from the -episode=N option
        #
        $filename = sprintf( $template, $showno );
    }
    elsif ( $optarg =~ /%d/ ) {
        #
        # There's an argument, does it have a '%d' in it?
        #
        $count = () = $optarg =~ /%d/g;
        die "Invalid - too many '%d' sequences in '$optarg'\n"
            if ( $count > 1 );
        $filename = sprintf( $optarg, $showno );
    }
    else {
        #
        # It's a plain filename, just return it
        #
        $filename = $optarg;
    }

    return $filename;
}

#===  FUNCTION  ================================================================
#         NAME: update_status
#      PURPOSE: Updates the status file
#   PARAMETERS: $sfile          Name and path of the status file
#               $state          State to set, such as 'parsed'
#      RETURNS: Nothing
#  DESCRIPTION: Appends the state string to the named file.
#       THROWS: No exceptions
#     COMMENTS: None
#     SEE ALSO: N/A
#===============================================================================
sub update_status {
    my ($sfile, $state) = @_;

    open( my $sfh, '>>:encoding(UTF-8)', $sfile )
        or die "Unable to open output file $sfile $!\n";
    print $sfh "$state\n";
    close($sfh);

}

#===  FUNCTION  ================================================================
#         NAME: spellcheck
#      PURPOSE: Perform simple spell checks on strings
#   PARAMETERS: $string         Input string to check
#      RETURNS: List containing any problem words or undef if none
#  DESCRIPTION: 
#       THROWS: No exceptions
#     COMMENTS: None
#     SEE ALSO: N/A
#===============================================================================
sub spellcheck {
    my ($string) = @_;

    my @errors;

    my $checker = Text::SpellChecker->new(
        text => $string,
        lang => 'en_GB.UTF-8'
    );

    while ( my $word = $checker->next_word ) {
        push( @errors, $word );
    }

    return @errors;

}

#===  FUNCTION  ================================================================
#         NAME: textFormat
#      PURPOSE: Formats a block of text in an indented, wrapped style with
#               a label in the left column
#   PARAMETERS: $text           The text to be formatted, as a scalar string
#               $tag            The label to be added to the left of the top
#                               line
#               $align          Tag alignment, 'L' for left, otherwise right
#               $lmargin        Width of the left margin (assumed to be big
#                               enough for the tag)
#               $textwidth      The width of all of the text plus left margin
#                               (i.e.  the right margin)
#      RETURNS: The formatted result as a string
#  DESCRIPTION: Chops the incoming text into words (thereby removing any
#               formatting). Removes any leading spaces. Loops through the
#               wordlist building them into lines of the right length to fit
#               between the left and right margins. Saves the lines in an
#               array. Adds the tag to the first line with the alignment
#               requested then returns the array joined into a string.
#       THROWS: No exceptions
#     COMMENTS: Inspired by Text::Format but *much* simpler. In fact T::F is
#               a nasty thing to have to use; I couldn't get it to do what
#               this routine does.
#               TODO Make the routine more resilient to silly input values.
#     SEE ALSO:
#===============================================================================
sub textFormat {
    my ( $text, $tag, $align, $lmargin, $textwidth ) = @_;

    my ( $width, $word );
    my ( @words, @buff, @wrap );

    #
    # Build the tag early. If there's no text we'll just return the tag.
    #
    $tag = sprintf( "%*s",
        ( $align =~ /L/i ? ( $lmargin - 1 ) * -1 : $lmargin - 1 ), $tag );

    return $tag unless $text;

    $text =~ s/(^\s+|\s+$)//g;
    return $tag unless $text;

    #
    # Chop up the incoming text removing leading spaces
    #
    @words = split( /\s+/, $text );
    shift(@words) if ( @words && $words[0] eq '' );

    #
    # Compute the width of the active text
    #
    $width = $textwidth - $lmargin;

    #
    # Format the words into lines with a blank left margin
    #
    while ( defined( $word = shift(@words) ) ) {
        if ( length( join( ' ', @buff, $word ) ) < $width ) {
            push( @buff, $word );
        }
        else {
            push( @wrap, ' ' x $lmargin . join( ' ', @buff ) );
            @buff = ($word);
        }
    }

    #
    # Append any remainder
    #
    push( @wrap, ' ' x $lmargin . join( ' ', @buff ) ) if @buff;

    #
    # Insert the tag into the first line
    #
    substr( $wrap[0], 0, $lmargin - 1 ) = $tag;

    #
    # Return the formatted array as a string
    #
    return join( "\n", @wrap );

}

#===  FUNCTION  ================================================================
#         NAME: colour_if
#      PURPOSE: Colours a string if a condition is true
#   PARAMETERS: $test           criterion to determine colouring
#               $colour         colour code to use
#               $string         string to colour
#      RETURNS: The (non-)coloured string
#  DESCRIPTION:
#       THROWS: No exceptions
#     COMMENTS: None
#     SEE ALSO: N/A
#===============================================================================
sub colour_if {
    my ( $test, $colour, $string ) = @_;

    if ($test) {
        return colored( $string, $colour );
    }
    else {
        return $string;
    }
}

#===  FUNCTION  ================================================================
#         NAME: colour_switch
#      PURPOSE: Chooses colours depending on multiple tests
#   PARAMETERS: $tests          arrayref containing a list of tests
#               $colours        arrayref containing a list of colour codes
#               $string         string to colour
#      RETURNS: The string with the desired colour (or none)
#  DESCRIPTION: The first test in @$tests which returns True causes the
#               corresponding colour to be chosen (unless it's missing)
#               otherwise, if no tests are True or the colour is missing,
#               there's no colour applied.
#       THROWS: No exceptions
#     COMMENTS: None
#     SEE ALSO: N/A
#===============================================================================
sub colour_switch {
    my ( $tests, $colours, $string ) = @_;

    for ( my $i = 0; $i <= $#{$tests}; $i++ ) {
        if ( $tests->[$i] ) {
            if ( exists( $colours->[$i] ) ) {
                return colored( $string, $colours->[$i] );
            }
            else {
                return $string;
            }
        }
    }

    return $string;
}

#===  FUNCTION  ================================================================
#         NAME: alert
#      PURPOSE: Print an alert message using a format and colours
#   PARAMETERS: $fmt            the format to use. The first field is where
#                               the string 'ALERT' goes
#               $message        the alert message
#      RETURNS: Nothing
#  DESCRIPTION:
#       THROWS: No exceptions
#     COMMENTS: None
#     SEE ALSO: N/A
#===============================================================================
sub alert {
    my ( $fmt, $message ) = @_;

    $fmt = "%-16s %s\n" unless $fmt;
    print STDERR colored( sprintf( $fmt, "** ALERT **:", $message ),
        'bold red' );
}

#===  FUNCTION  ================================================================
#         NAME: coalesce
#      PURPOSE: To find the first defined argument and return it
#   PARAMETERS: Arbitrary number of arguments
#      RETURNS: The first defined argument or undef if there are none
#  DESCRIPTION:
#       THROWS: No exceptions
#     COMMENTS: None
#     SEE ALSO: N/A
#===============================================================================
sub coalesce {
    foreach (@_) {
        return $_ if defined($_);
    }
    return undef;    ## no critic
}

#===  FUNCTION  ================================================================
#         NAME: quote_crlf
#      PURPOSE: Display TAB (\t) and CRLF (\r\n) sequences for debugging
#   PARAMETERS: $string         String to process
#      RETURNS: The string with TAB, CR and LF characters made visible for
#               printing
#  DESCRIPTION:
#       THROWS: No exceptions
#     COMMENTS: None
#     SEE ALSO: N/A
#===============================================================================
sub quote_crlf {
    my ($string) = @_;

    my $quoted = $string;
    $quoted =~ s/\t/\\t/g;
    $quoted =~ s/\n/\\n/g;
    $quoted =~ s/\r/\\r/g;

    return "$quoted\n";
}

#===  FUNCTION  ================================================================
#         NAME: _debug
#      PURPOSE: Prints debug reports
#   PARAMETERS: $active         Boolean: 1 for print, 0 for no print
#               $message        Message to print
#      RETURNS: Nothing
#  DESCRIPTION: Outputs a message if $active is true. It removes any trailing
#               newline and then adds one in the 'print' to the caller doesn't
#               have to bother. Prepends the message with 'D> ' to show it's
#               a debug message.
#       THROWS: No exceptions
#     COMMENTS: None
#     SEE ALSO: N/A
#===============================================================================
sub _debug {
    my ( $active, $message ) = @_;

    chomp($message);
    print STDERR "D> $message\n" if $active;
}

#===  FUNCTION  ================================================================
#         NAME: Options
#      PURPOSE: Processes command-line options
#   PARAMETERS: $optref     Hash reference to hold the options
#      RETURNS: Undef
#  DESCRIPTION:
#       THROWS: no exceptions
#     COMMENTS: none
#     SEE ALSO: n/a
#===============================================================================
sub Options {
    my ($optref) = @_;

    my @options = (
        "help",        "documentation|manpage",
        "debug=i",     "silent!",
        "episode=i",   "infile=s",
        "shownotes:s", "format=s",
        "release=s",   "assets=s",
        "pictures=s",  "zip=s",
        "test!",
    );

    if ( !GetOptions( $optref, @options ) ) {
        pod2usage(
            -msg     => "$PROG version $VERSION\n",
            -verbose => 0,
            -exitval => 1
        );
    }

    return;
}

__END__


#%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
#  Application Documentation
#%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
#{{{

=head1 NAME

parse_JSON - parse the JSON output file from the HPR show submission form

=head1 VERSION

This documentation refers to parse_JSON version 0.0.17


=head1 USAGE

    ./parse_JSON [-help] [-documentation|manpage] -episode=N -infile=FILE
         [-shownotes[=FILE]] [-format=FILE] [-release=FILE] [-assets=FILE]
         [-pictures=FILE] [-zip=FILE] [-debug=N] [-[no]silent]


    ./parse_JSON -episode=2384 -infile=shownotes/hpr2384/shownotes.json
    ./parse_JSON -episode=2379 -infile=shownotes/hpr2379/shownotes.json -shownotes


=head1 OPTIONS

=over 8

=item B<-help>

Prints a brief help message describing the usage of the program, and then exits.

=item B<-manpage>

Prints the entire documentation for the script. Note that this mode should use
I<perldoc> but there seems to be a bug in this tool at the moment; it does not
render properly. Currently we are using a simpler type of formatter.

=item B<-episode=N>

This mandatory option specifies the number of the HPR episode that is being
parsed.

=item B<-infile=FILE>

This mandatory option provides the file containing the JSON-format output from
the submission form on the HPR website. It is expected that the file has been
collected by external scripts and saved in a local file.

=item B<-shownotes[=FILE]>

This option, which may be omitted, defines the location where the parsed show
notes are to be written. If omitted no show notes are written, the script
simply parses the input file and reports on what it finds.

If the option is given as B<-shownotes=FILE> the notes are written to the
nominated file.

The B<FILE> portion may be given as a template string containing a I<%d>
sequence. This sequence is replaced by the show number provided by the
B<-episode=N> option.

If the B<=FILE> portion is omitted a default template of 'hpr%d.out' is used
and again the I<%d> sequence is replaced by the show number provided by the
B<-episode=N> option.

If the notes are declared as HTML5 then a symbolic link is made from the
parsed shownote file to a file where the extension is 'html'. This is to allow
the 'do_vim' script to edit the output file as in other cases but to ensure
that the file to be uploaded to the server is also edited.

=item B<-format=FILE>

This option, which may be omitted, causes the declared format to be written to
a nominated file. This is to allow other scripts to take action depending on
the format of the notes without having to determine this information
themselves.

If the option is omitted no format information is written.

=item B<-release=FILE>

This option, which may be omitted, causes the show's release date to be
written to a nominated file. This is to allow other scripts to take action
depending on the release date of the show without having to determine this
information themselves.

If the option is omitted no release date is written.

=item B<-assets=FILE>

This option, which may be omitted, causes a list of the show's assets to be
written to a nominated file. This is to allow other scripts to take action
depending on the assets of the show without having to determine this
information themselves.

Note that the assets do not include the main audio file.

If no assets exist or the option is omitted no asset list is written.

=item B<-pictures=FILE>

This option, which may be omitted, causes a list of any picture assets to be
written to a nominated file. This is to allow other scripts to take action
depending on pictures sent with the show without having to determine this
information themselves.

If no pictures exist or the option is omitted no picture list is written.

=item B<-zip=FILE>

This option, which may be omitted, causes assets to written to a Zip file
with the name given. This only happens if the assets were sent separately, and
not in an archive file of some kind. This is because the processing of some
assets (only pictures at the moment) can cause changes to be made to the
originals. This step can be taken to preserve the original files.

If no assets exist or an archive file already exists, or the option is omitted
no zip file is written.

=item B<-debug=N>

Causes certain debugging information to be displayed.

    0   (the default) no debug output
    1   N/A
    2   reports lines being parsed from the input file
    3   dumps two data structures:
            - @content: an array containing lines from the input file
            - %showdetails: a hash containing the parsed data

=item B<-[no]silent>

Controls the amount of information generated. If B<-silent> is selected the
output is reduced, otherwise B<-nosilent> is assumed or may be provided when
information about the generation of output files is reported.

=item B<-[no]test>

Used when running tests on the script. Controls the log output, adding 'T ' at
the start of the log record. If B<-test> is selected the
prefix is added, otherwise B<-notest> is assumed and no prefix is added.

=back

=head1 DESCRIPTION

This script is meant to be a component in a workflow dealing with the notes
submitted by HPR hosts with the shows they are uploading.

The script reads a JSON file generated by the show upload form on the HPR
website.  It decodes this into a hash structure containing all of the elements
of the file.

Regardless of settings a check is made to ensure that the episode number
provided to the script matches that found in the input file.

An analysis of the show notes is carried out to determine how much HTML they
contain. The HTML level is determined as a percentage. If it is above 5% then
the notes are deemed to be HTML. If below 5% then an attempt is made to
determine what type of markup has been used.

An analysis of the host details is also made to determine the composition of
the text therein.

Reports are generated for all of these components. A recommendation of the
type of processing the notes will require is also made. This was intended to
assist with automatic processing, but has not proved to be reliable enough.

The author of the notes declares what format they are. This is not currently
compared with what the analysis reports, but this may be added in a later
version.

The notes are written to the nominated file (if there is one) for further
processing, database insertion, etc.

If the notes are declared to be HTML then a copy is written to a file with the
same name as the output file but with an 'html' suffix. This can be rendered
to see if the notes really are HTML and look good enough for uploading.


=head1 DIAGNOSTICS

Error and warning messages generated by the script.

=over 4

=item B<Unable to access input file ...: ...>

Type: fatal

The script attempted to open the input file but failed. The message will
contain details of the failure.

=item B<Input file ... is empty>

Type: fatal

The input file exists but is empty

=item B<Invalid - too many '%d' sequences in '...'>

Type: fatal

The template provided to the B<-shownotes=FILE> option contained too many '%d'
sequences. Ensure that there is only one of these. See above for details.

=item B<The JSON doesn't have the right structure>

Type: fatal

After a few basic checks on the components of the JSON file it does not seem
to be valid. The script expects there at least to be top-level sections with
the keys: I<episode>, I<metadata> and I<host>.

=item B<The episode specified (...) doesn't match the one in the JSON (...)>

Type: fatal

The episode number in the B<-episode=N> option does not match the episode
number found in the JSON. Both numbers are reported in the error message.

=item B<Unable to open output file ...: ...>

Type: fatal

The script was unable to open one or both of the output files or the file of
format information. These will be named according to the B<-shownotes=FILE>
option with a suffix of '.out' or '.html' or specified by the B<-format=FILE>
option.

=back

=head1 DEPENDENCIES

    Data::Dumper
    Getopt::Long
    HTML::Parser
    JSON
    Link::Extor
    Log::Handler
    Pod::Usage
    Term::ANSIColor
    Text::CSV

=head1 BUGS AND LIMITATIONS

There are no known bugs in this module.
Please report problems to Dave Morriss  (Dave.Morriss@gmail.com)
Patches are welcome.

=head1 AUTHOR

Dave Morriss (Dave.Morriss@gmail.com)

=head1 LICENCE AND COPYRIGHT

Copyright (c) 2021 Dave Morriss (Dave.Morriss@gmail.com). All rights reserved.

This module is free software; you can redistribute it and/or
modify it under the same terms as Perl itself. See perldoc perlartistic.

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.

=cut

#}}}

# [zo to open fold, zc to close]

# vim: syntax=perl:ts=8:sw=4:et:ai:tw=78:fo=tcrqn21:fdm=marker