Gecko [ fernandoquevedodop.com ]

Name	Size	Permission
Charset	[ DIR ]	drwxr-xr-x
Lang	[ DIR ]	drwxr-xr-x
AES.pm	18.66 KB	-rw-r--r--
AFCP.pm	10.38 KB	-rw-r--r--
AIFF.pm	9.59 KB	-rw-r--r--
APE.pm	9.06 KB	-rw-r--r--
APP12.pm	10.44 KB	-rw-r--r--
ASF.pm	30.21 KB	-rw-r--r--
Apple.pm	5.43 KB	-rw-r--r--
Audible.pm	11.52 KB	-rw-r--r--
BMP.pm	10.45 KB	-rw-r--r--
BPG.pm	7.53 KB	-rw-r--r--
BZZ.pm	17.46 KB	-rw-r--r--
BigTIFF.pm	11.27 KB	-rw-r--r--
BuildTagLookup.pm	121.94 KB	-rw-r--r--
CBOR.pm	11.51 KB	-rw-r--r--
Canon.pm	375.39 KB	-rw-r--r--
CanonCustom.pm	84.32 KB	-rw-r--r--
CanonRaw.pm	29.68 KB	-rw-r--r--
CanonVRD.pm	75.2 KB	-rw-r--r--
CaptureOne.pm	7.85 KB	-rw-r--r--
Casio.pm	62.21 KB	-rw-r--r--
Charset.pm	16.01 KB	-rw-r--r--
DICOM.pm	245.09 KB	-rw-r--r--
DJI.pm	6.31 KB	-rw-r--r--
DNG.pm	31.59 KB	-rw-r--r--
DPX.pm	9.51 KB	-rw-r--r--
DV.pm	9.75 KB	-rw-r--r--
DarwinCore.pm	15.43 KB	-rw-r--r--
DjVu.pm	12.46 KB	-rw-r--r--
EXE.pm	50.3 KB	-rw-r--r--
Exif.pm	246.64 KB	-rw-r--r--
FITS.pm	5.73 KB	-rw-r--r--
FLAC.pm	9.89 KB	-rw-r--r--
FLIF.pm	11.65 KB	-rw-r--r--
FLIR.pm	59.45 KB	-rw-r--r--
Fixup.pm	12.43 KB	-rw-r--r--
Flash.pm	27.5 KB	-rw-r--r--
FlashPix.pm	87.44 KB	-rw-r--r--
Font.pm	27.34 KB	-rw-r--r--
FotoStation.pm	7.97 KB	-rw-r--r--
FujiFilm.pm	54.47 KB	-rw-r--r--
GE.pm	2 KB	-rw-r--r--
GIF.pm	21.85 KB	-rw-r--r--
GIMP.pm	8.35 KB	-rw-r--r--
GPS.pm	19.53 KB	-rw-r--r--
GeoTiff.pm	87.51 KB	-rw-r--r--
Geotag.pm	62.87 KB	-rw-r--r--
GoPro.pm	28 KB	-rw-r--r--
H264.pm	39.29 KB	-rw-r--r--
HP.pm	8 KB	-rw-r--r--
HTML.pm	22.1 KB	-rw-r--r--
HtmlDump.pm	32.8 KB	-rw-r--r--
ICC_Profile.pm	45.63 KB	-rw-r--r--
ID3.pm	56.91 KB	-rw-r--r--
IPTC.pm	39.54 KB	-rw-r--r--
ISO.pm	6.81 KB	-rw-r--r--
ITC.pm	6.63 KB	-rw-r--r--
Import.pm	12.83 KB	-rw-r--r--
InDesign.pm	10.4 KB	-rw-r--r--
JPEG.pm	24.3 KB	-rw-r--r--
JPEGDigest.pm	212.38 KB	-rw-r--r--
JSON.pm	6.15 KB	-rw-r--r--
JVC.pm	3.57 KB	-rw-r--r--
Jpeg2000.pm	49.65 KB	-rw-r--r--
Kodak.pm	120.47 KB	-rw-r--r--
KyoceraRaw.pm	4.31 KB	-rw-r--r--
LIF.pm	4.66 KB	-rw-r--r--
LNK.pm	22.02 KB	-rw-r--r--
Leaf.pm	16.34 KB	-rw-r--r--
Lytro.pm	7.56 KB	-rw-r--r--
M2TS.pm	37.11 KB	-rw-r--r--
MIE.pm	102.09 KB	-rw-r--r--
MIEUnits.pod	13.68 KB	-rw-r--r--
MIFF.pm	9.05 KB	-rw-r--r--
MNG.pm	16.22 KB	-rw-r--r--
MOI.pm	4.14 KB	-rw-r--r--
MPC.pm	4.05 KB	-rw-r--r--
MPEG.pm	22.32 KB	-rw-r--r--
MPF.pm	8.94 KB	-rw-r--r--
MRC.pm	18.39 KB	-rw-r--r--
MWG.pm	30.64 KB	-rw-r--r--
MXF.pm	252.65 KB	-rw-r--r--
MacOS.pm	30.7 KB	-rw-r--r--
MakerNotes.pm	68.93 KB	-rw-r--r--
Matroska.pm	34.44 KB	-rw-r--r--
Microsoft.pm	52.13 KB	-rw-r--r--
Minolta.pm	99.93 KB	-rw-r--r--
MinoltaRaw.pm	16.49 KB	-rw-r--r--
Motorola.pm	5.31 KB	-rw-r--r--
Nikon.pm	431.96 KB	-rw-r--r--
NikonCapture.pm	28.13 KB	-rw-r--r--
NikonCustom.pm	290.78 KB	-rw-r--r--
NikonSettings.pm	71.88 KB	-rw-r--r--
Nintendo.pm	3.24 KB	-rw-r--r--
OOXML.pm	13.61 KB	-rw-r--r--
Ogg.pm	8.81 KB	-rw-r--r--
Olympus.pm	139.89 KB	-rw-r--r--
OpenEXR.pm	10.26 KB	-rw-r--r--
Opus.pm	2.22 KB	-rw-r--r--
Other.pm	2.63 KB	-rw-r--r--
PCX.pm	3.72 KB	-rw-r--r--
PDF.pm	90.77 KB	-rw-r--r--
PGF.pm	4.04 KB	-rw-r--r--
PICT.pm	33.2 KB	-rw-r--r--
PLIST.pm	17.07 KB	-rw-r--r--
PLUS.pm	147.71 KB	-rw-r--r--
PNG.pm	60.72 KB	-rw-r--r--
PPM.pm	4.97 KB	-rw-r--r--
PSP.pm	9.02 KB	-rw-r--r--
Palm.pm	11.9 KB	-rw-r--r--
Panasonic.pm	96.62 KB	-rw-r--r--
PanasonicRaw.pm	31.77 KB	-rw-r--r--
Parrot.pm	20.95 KB	-rw-r--r--
Pentax.pm	220.46 KB	-rw-r--r--
PhaseOne.pm	26.24 KB	-rw-r--r--
PhotoCD.pm	18.25 KB	-rw-r--r--
PhotoMechanic.pm	8.09 KB	-rw-r--r--
Photoshop.pm	43.36 KB	-rw-r--r--
PostScript.pm	29.11 KB	-rw-r--r--
PrintIM.pm	3.44 KB	-rw-r--r--
Qualcomm.pm	43.43 KB	-rw-r--r--
QuickTime.pm	409.72 KB	-rw-r--r--
QuickTimeStream.pl	126.03 KB	-rw-r--r--
RIFF.pm	70.97 KB	-rw-r--r--
RSRC.pm	9.58 KB	-rw-r--r--
RTF.pm	13.46 KB	-rw-r--r--
Radiance.pm	4.56 KB	-rw-r--r--
Rawzor.pm	6.17 KB	-rw-r--r--
Real.pm	26.89 KB	-rw-r--r--
Reconyx.pm	14.2 KB	-rw-r--r--
Red.pm	10.82 KB	-rw-r--r--
Ricoh.pm	37.18 KB	-rw-r--r--
Samsung.pm	54.64 KB	-rw-r--r--
Sanyo.pm	11.83 KB	-rw-r--r--
Scalado.pm	3.91 KB	-rw-r--r--
Shift.pl	23.28 KB	-rw-r--r--
Shortcuts.pm	9.81 KB	-rw-r--r--
Sigma.pm	33.16 KB	-rw-r--r--
SigmaRaw.pm	23.45 KB	-rw-r--r--
Sony.pm	439.69 KB	-rw-r--r--
SonyIDC.pm	11.95 KB	-rw-r--r--
Stim.pm	4.53 KB	-rw-r--r--
TagInfoXML.pm	33.14 KB	-rw-r--r--
TagLookup.pm	647.18 KB	-rw-r--r--
TagNames.pod	1.76 MB	-rw-r--r--
Text.pm	8.12 KB	-rw-r--r--
Theora.pm	3.41 KB	-rw-r--r--
Torrent.pm	11.58 KB	-rw-r--r--
Unknown.pm	1.69 KB	-rw-r--r--
VCard.pm	17.31 KB	-rw-r--r--
Validate.pm	29.75 KB	-rw-r--r--
Vorbis.pm	8.13 KB	-rw-r--r--
WTV.pm	12.97 KB	-rw-r--r--
WriteCanonRaw.pl	24.29 KB	-rw-r--r--
WriteExif.pl	130.6 KB	-rw-r--r--
WriteIPTC.pl	29.6 KB	-rw-r--r--
WritePDF.pl	29.02 KB	-rw-r--r--
WritePNG.pl	16.07 KB	-rw-r--r--
WritePhotoshop.pl	10.85 KB	-rw-r--r--
WritePostScript.pl	26.71 KB	-rw-r--r--
WriteQuickTime.pl	89.63 KB	-rw-r--r--
WriteXMP.pl	70.49 KB	-rw-r--r--
Writer.pl	301.76 KB	-rw-r--r--
XMP.pm	176.89 KB	-rw-r--r--
XMP2.pl	81.66 KB	-rw-r--r--
XMPStruct.pl	36.66 KB	-rw-r--r--
ZIP.pm	27.61 KB	-rw-r--r--
ZISRAW.pm	7.85 KB	-rw-r--r--
iWork.pm	7.79 KB	-rw-r--r--

Code Editor : Text.pm

#------------------------------------------------------------------------------
# File:         Text.pm
#
# Description:  Deduce characteristics of TXT and CSV files
#
# Revisions:    2019-11-01 - P. Harvey Created
#               2020-02-13 - PH Added CSV file support
#
# References:   1) https://github.com/file/file
#------------------------------------------------------------------------------

package Image::ExifTool::Text;

use strict;
use vars qw($VERSION);
use Image::ExifTool qw(:DataAccess :Utils);
use Image::ExifTool::XMP;

$VERSION = '1.03';

# Text tags
%Image::ExifTool::Text::Main = (
    VARS => { NO_ID => 1 },
    GROUPS => { 0 => 'File', 1 => 'File', 2 => 'Document' },
    NOTES => q{
        Although basic text files contain no metadata, the following tags are
        determined from a simple analysis of the data in TXT and CSV files. 
        Statistics are generated only for 8-bit encodings, but the L<FastScan|../ExifTool.html#FastScan> (-fast)
        option may be used to limit processing to the first 64 kB in which case some
        tags are not produced.  To avoid long processing delays, ExifTool will issue
        a minor warning and process only the first 64 kB of any file larger than 20
        MB unless the L<IgnoreMinorErrors|../ExifTool.html#IgnoreMinorErrors> (-m)
        option is used.
    },
    MIMEEncoding => { Groups => { 2 => 'Other' } },
    Newlines => {
        PrintConv => {
            "\r\n" => 'Windows CRLF',
            "\r"   => 'Macintosh CR',
            "\n"   => 'Unix LF',
            ''     => '(none)',
        },
    },
    ByteOrderMark => { PrintConv => { 0 => 'No', 1 => 'Yes' } },
    LineCount => { },
    WordCount => { },
    Delimiter => { PrintConv => { '' => '(none)', ',' => 'Comma', ';' => 'Semicolon', "\t" => 'Tab' }},
    Quoting   => { PrintConv => { '' => '(none)', '"' => 'Double quotes', "'" => 'Single quotes' }},
    RowCount  => { },
    ColumnCount => { },
);

#------------------------------------------------------------------------------
# Extract some stats from a text file
# Inputs: 0) ExifTool ref, 1) dirInfo ref
# Returns: 1 on success, 0 if this wasn't a text file
sub ProcessTXT($$)
{
    my ($et, $dirInfo) = @_;
    my $dataPt = $$dirInfo{TestBuff};
    my $raf = $$dirInfo{RAF};
    my $fast = $et->Options('FastScan') || 0;
    my ($buff, $enc, $isBOM, $isUTF8);
    my $nl = '';

return 0 unless length $$dataPt; # can't call it a text file if it has no text

# read more from the file if necessary
    if ($fast < 3 and length($$dataPt) == $Image::ExifTool::testLen) {
        $raf->Read($buff, 65536) or return 0;
        $dataPt = \$buff;
    }
#
# make our best guess at the character encoding (EBCDIC is not supported)
#
    if ($$dataPt =~ /([\0-\x06\x0e-\x1a\x1c-\x1f\x7f])/) {
        # file contains weird control characters, could be multi-byte Unicode
        if ($$dataPt =~ /^(\xff\xfe\0\0|\0\0\xfe\xff)/) {
            if ($1 eq "\xff\xfe\0\0") {
                $enc = 'utf-32le';
                $nl = $1 if $$dataPt =~ /(\r\0\0\0\n|\r|\n)\0\0\0/;
            } else {
                $enc = 'utf-32be';
                $nl = $1 if $$dataPt =~ /\0\0\0(\r\0\0\0\n|\r|\n)/;
            }
        } elsif ($$dataPt =~ /^(\xff\xfe|\xfe\xff)/) {
            if ($1 eq "\xff\xfe") {
                $enc = 'utf-16le';
                $nl = $1 if $$dataPt =~ /(\r\0\n|\r|\n)\0/;
            } else {
                $enc = 'utf-16be';
                $nl = $1 if $$dataPt =~ /\0(\r\0\n|\r|\n)/;
            }
        } else {
            return 0;       # probably not a text file
        }
        $nl =~ tr/\0//d;    # remove nulls from newline sequence
        $isBOM = 1;         # (we don't recognize UTF-16/UTF-32 without one)
    } else {
        $isUTF8 = Image::ExifTool::XMP::IsUTF8($dataPt, 1);
        if ($isUTF8 == 0) {
            $enc = 'us-ascii';
        } elsif ($isUTF8 > 0) {
            $enc = 'utf-8';
            $isBOM = ($$dataPt =~ /^\xef\xbb\xbf/ ? 1 : 0);
        } elsif ($$dataPt !~ /[\x80-\x9f]/) {
            $enc = 'iso-8859-1';
        } else {
            $enc = 'unknown-8bit';
        }
        $nl = $1 if $$dataPt =~ /(\r\n|\r|\n)/;
    }

my $tagTablePtr = GetTagTable('Image::ExifTool::Text::Main');

$et->SetFileType();
    $et->HandleTag($tagTablePtr, MIMEEncoding => $enc);

return 1 if $fast == 3 or not $raf->Seek(0,0);

$et->HandleTag($tagTablePtr, ByteOrderMark => $isBOM) if defined $isBOM;
    $et->HandleTag($tagTablePtr, Newlines => $nl);

return 1 if $fast or not defined $isUTF8;
#
# generate stats for CSV files
#
    if ($$et{FileType} eq 'CSV') {
        my ($delim, $quot, $ncols);
        my $nrows = 0;
        while ($raf->ReadLine($buff)) {
            if (not defined $delim) {
                my %count = ( ',' => 0, ';' => 0, "\t" => 0 );
                ++$count{$_} foreach $buff =~ /[,;\t]/g;
                if ($count{','} > $count{';'} and $count{','} > $count{"\t"}) {
                    $delim = ',';
                } elsif ($count{';'} > $count{"\t"}) {
                    $delim = ';';
                } elsif ($count{"\t"}) {
                    $delim = "\t";
                } else {
                    $delim = '';
                    $ncols = 1;
                }
                unless ($ncols) {
                    # account for delimiters in quotes (simplistically)
                    while ($buff =~ /(^|$delim)(["'])(.*?)\2(?=$delim|$)/sg) {
                        $quot = $2;
                        my $field = $3;
                        $count{$delim} -= () = $field =~ /$delim/g;
                    }
                    $ncols = $count{$delim} + 1;
                }
            } elsif (not $quot) {
                $quot = $2 if $buff =~ /(^|$delim)(["'])(.*?)\2(?=$delim|$)/sg;
            }
            if (++$nrows == 1000 and $et->Warn('Not counting rows past 1000', 2)) {
                undef $nrows;
                last;
            }
        }
        $et->HandleTag($tagTablePtr, Delimiter => ($delim || ''));
        $et->HandleTag($tagTablePtr, Quoting => ($quot || ''));
        $et->HandleTag($tagTablePtr, ColumnCount => $ncols);
        $et->HandleTag($tagTablePtr, RowCount => $nrows) if $nrows;
        return 1;
    }
    return 1 if $$et{VALUE}{FileSize} and $$et{VALUE}{FileSize} > 20000000 and
        $et->Warn('Not counting lines/words in text file larger than 20 MB', 2);
#
# count lines/words and check encoding of the rest of the file
#
    my ($lines, $words) = (0, 0);
    my $oldNL = $/;
    $/ = $nl if $nl;
    while ($raf->ReadLine($buff)) {
        ++$lines;
        ++$words while $buff =~ /\S+/g;
        if (not $nl and $buff =~ /(\r\n|\r|\n)$/) {
            # (the first line must have been longer than 64 kB)
            $$et{VALUE}{Newlines} = $nl = $1;
        }
        next if $raf->Tell() < 65536;
        # continue to check encoding after the first 64 kB
        if ($isUTF8 >= 0) { # (if ascii or utf8)
            $isUTF8 = Image::ExifTool::XMP::IsUTF8(\$buff);
            if ($isUTF8 > 0) {
                $enc = 'utf-8';
            } elsif ($isUTF8 < 0) {
                $enc = $buff =~ /[\x80-\x9f]/ ? 'unknown-8bit' : 'iso-8859-1';
            }
        } elsif ($enc eq 'iso-8859-1' and $buff =~ /[\x80-\x9f]/) {
            $enc = 'unknown-8bit';
        }
    }
    if ($$et{VALUE}{MIMEEncoding} ne $enc) {
        $$et{VALUE}{MIMEEncoding} = $enc;
        $et->VPrint(0,"  MIMEEncoding [override] = $enc\n");
    }
    $/ = $oldNL;
    $et->HandleTag($tagTablePtr, LineCount => $lines);
    $et->HandleTag($tagTablePtr, WordCount => $words);
    return 1;
}

1;  # end

__END__

=head1 NAME

Image::ExifTool::Text - Read Text meta information

=head1 SYNOPSIS

This module is used by Image::ExifTool

=head1 DESCRIPTION

This module contains definitions required by Image::ExifTool to deduce some
characteristics of TXT and CSV files.

=head1 AUTHOR

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

=head1 REFERENCES

=over 4

=item L<https://github.com/file/file>

=back

=head1 SEE ALSO

L<Image::ExifTool::TagNames/Text Tags>,
L<Image::ExifTool(3pm)|Image::ExifTool>

=cut

${this.title}

Code Editor : Text.pm