exiftool_vendored 12.09.0 → 12.14.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 1095c3210c7799b406dbff7b262f1e48bba57874a51d7ada486b7ab12c4f946c
-  data.tar.gz: 6cd2472d73f0bd3178bfb9cb31d3c2dbd3c37d5c73ba370ea6fced3a9bfde512
+  metadata.gz: 1529a49f0175445c31e6d5a6607546d958e2b08ce2c044d43232ac07f51a6fb5
+  data.tar.gz: bc6637fc0ef323192aa5e6c7e8553520f2d0acd9039e870ae1f9340fa00445c8
 SHA512:
-  metadata.gz: 490b8624766262c4231eb655f2df5e6c2ebe35efa3f4e8131b36696d017daf5d637e3baad7fb403c587ab18e2fe09321a8444bd797eb093c127bdaabbd2f38ed
-  data.tar.gz: fd4dd627988588ee71fee72a7071645aa9ef53bb687e0ade23e53c0e63590912f81abf6a300bf969e2578a64a0f58d962417d0dae00779c0ff7ea4987c5af28e
+  metadata.gz: 03fa438994202e58a6c85ae0715018f89ca76c3f0fbfa87a9a8438b5c73637ed53d92491759f63926e3c57cefb2ec5f22dadd8b67c04d20a8e06aaa33f0328c0
+  data.tar.gz: 5e2cd8a33c4d66057bf115cd6f3bf3adaab426a44d8988747581d21478c81325e3cc88410b783cb6da7a55efbc04f7bc448291d509baf248995d10cfdcdeea0f

data/bin/Changes

@@ -5,7 +5,80 @@ ExifTool Version History
 RSS feed: https://exiftool.org/rss.xml
 
 Note: The most recent production release is Version 12.00. (Other versions are
-considered development releases, and are not uploaded to CPAN.)
+considered development releases, and are not uploaded to MetaCPAN.)
+
+Jan. 6, 2021 - Version 12.14
+
+  - Added support for 2 more types of timed GPS in video files (that makes 49
+    different formats now supported)
+  - Added validity check for PDF trailer dictionary Size
+  - Added a new Pentax LensType
+  - Extract metadata from Jpeg2000 Association box
+  - Changed -g:XX:YY and -G:XX:YY options to show empty strings for non-existent
+    groups
+  - Patched to issue warning and avoid writing date/time values with a zero
+    month or day number
+  - Patched to avoid runtime warnings if trying to set FileName to an empty
+    string
+  - Fixed issue that could cause GPS test number 12 to fail on some systems
+  - Fixed problem extracting XML as a block from Jpeg2000 images, and extract
+    XML tags in the XML group instead of XMP
+
+Dec. 24, 2020 - Version 12.13
+
+  - Added -i HIDDEN option to ignore files with names that start with "."
+  - Added a few new Nikon ShutterMode values (thanks Jan Skoda)
+  - Added ability to write Google GCamera MicroVideo XMP tags
+  - Add time zone automatically to most string-based QuickTime date/time tags
+    when writing unless the PrintConv option is disabled
+  - Decode a new Sony tag (thanks LibRaw)
+  - Changed behaviour when writing only pseudo tags to return an error and avoid
+    writing any other tags if writing FileName fails
+  - Print "X image files read" message even if only 1 file is read when at least
+    one other file has failed the -if condition
+
+Dec. 4, 2020 - Version 12.12
+
+  - Added ability to geotag from DJI CSV log files
+  - Added a new CanonModelID
+  - Added a couple of new Sony LensType values (thanks LibRaw)
+  - Enhanced -csvDelim option to allow "\t", "\n", "\r" and "\\"
+  - Unescape "\b" and "\f" in imported JSON values
+  - Fixed bug introduced in 12.10 which generated a "Not an integer" warning
+    when attempting to shift some QuickTime date/time tags
+  - Fixed shared-write permission problem with -@ argfile when using -stay_open
+    and a filename containing special characters on Windows
+
+Nov. 27, 2020 - Version 12.11
+
+  - Added -csvDelim option
+  - Added new Canon and Olympus LensType values (thanks LibRaw)
+  - Added a warning if ICC_Profile is deleted from an image (github issue #63)
+  - EndDir() function for -if option now works when -fileOrder is used
+  - Changed FileSize conversion to use binary prefixes since that is how the
+    conversion is currently done (eg. MiB instead of MB)
+  - Patched -csv option so columns aren't resorted when using -G option and one
+    of the tags is missing from a file
+  - Fixed incompatiblity with Google Photos when writing UserData:GPSCoordinates
+    to MP4 videos
+  - Fixed problem where the tags available in a -p format string were limited to
+    the same as the -if[NUM] option when NUM was specified
+  - Fixed incorrect decoding of SourceFileIndex/SourceDirectoryIndex for Ricoh
+    models
+
+Nov. 12, 2020 - Version 12.10
+
+  - Added -validate test for proper TIFF magic number in JPEG EXIF header
+  - Added support for Nikon Z7 LensData version 0801
+  - Added a new XMP-GPano tag
+  - Decode ColorData for the Canon EOS 1DXmkIII (thanks LibRaw)
+  - Decode more tags for the Sony ILCE-7SM3 (thanks Jos Roost)
+  - Automatically apply QuickTimeUTC option for CR3 files
+  - Improved decoding of XAttrMDLabel from MacOS files
+  - Ignore time zones when writing date/time values and using the -d option
+  - Enhanced -echo3 and -echo4 options to allow exit status to be returned
+  - Changed -execute so the -q option no longer suppresses the "{ready}" message
+    when a synchronization number is used (eg. -execute123)
 
 Oct. 29, 2020 - Version 12.09
 
@@ -17,15 +90,14 @@ Oct. 29, 2020 - Version 12.09
   - Added a new Sony LensType (thanks Jos Roost)
   - Added a new Nikon Z lens (thanks LibRaw)
   - Added a new Canon LensType
-  - Added a new XMP-crs tag
   - Decode ColorData for Canon EOS R5/R6
   - Decode a couple of new HEIF tags
   - Decode FirmwareVersion for Canon M50
-  - Patched EndDir() feature so subdirectories are always processed when -r is
-    used (previously, EndDir() would end processing of a directory completely)
   - Improved decoding of Sony CreativeStyle tags (thanks Jos Roost)
   - Improved parsing of Radiance files to recognize comments
   - Renamed GIF AspectRatio tag to PixelAspectRatio
+  - Patched EndDir() feature so subdirectories are always processed when -r is
+    used (previously, EndDir() would end processing of a directory completely)
   - Yet another tweak to the EventTime formatting rules (also allow time-only
     values with fractional seconds and a time zone)
   - Avoid loading GoPro module unnecessarily when reading MP4 videos from some
@@ -33,7 +105,6 @@ Oct. 29, 2020 - Version 12.09
   - Fixed problem with an incorrect naming of CodecID tags in some MKV videos
   - Fixed verbose output to avoid "adding" messages for existing flattened XMP
     tags
-  - Added read support for MacOS "._" sidecar files
 
 Oct. 15, 2020 - Version 12.08
 

data/bin/MANIFEST

@@ -567,6 +567,7 @@ t/GeoTiff_4.out
 t/Geotag.t
 t/Geotag_10.out
 t/Geotag_11.out
+t/Geotag_12.out
 t/Geotag_2.out
 t/Geotag_3.out
 t/Geotag_5.out
@@ -602,6 +603,7 @@ t/Jpeg2000.t
 t/Jpeg2000_2.out
 t/Jpeg2000_3.out
 t/Jpeg2000_4.out
+t/Jpeg2000_5.out
 t/Kodak.t
 t/Kodak_2.out
 t/Kodak_3.out
@@ -986,6 +988,7 @@ t/images/Geotag.log
 t/images/Geotag.xml
 t/images/Geotag2.log
 t/images/Geotag3.log
+t/images/Geotag_DJI_2020-12-02_[07-50-31].csv
 t/images/GoPro.jpg
 t/images/HTML.html
 t/images/ICC_Profile.icc

data/bin/META.json

@@ -47,5 +47,5 @@
       }
    },
    "release_status" : "stable",
-   "version" : "12.09"
+   "version" : "12.14"
 }

data/bin/META.yml

@@ -28,4 +28,4 @@ recommends:
   Time::HiRes: 0
 requires:
   perl: 5.004
-version: 12.09
+version: 12.14

data/bin/README

@@ -106,8 +106,8 @@ your home directory, then you would type the following commands in a
 terminal window to extract and run ExifTool:
 
   cd ~/Desktop
-  gzip -dc Image-ExifTool-12.09.tar.gz | tar -xf -
-  cd Image-ExifTool-12.09
+  gzip -dc Image-ExifTool-12.14.tar.gz | tar -xf -
+  cd Image-ExifTool-12.14
   ./exiftool t/images/ExifTool.jpg
 
 Note:  These commands extract meta information from one of the test images. 
@@ -161,7 +161,7 @@ for calculating digest values and providing other features listed below:
 
 COPYRIGHT AND LICENSE
 
-Copyright 2003-2020, Phil Harvey
+Copyright 2003-2021, Phil Harvey
 
 This is free software; you can redistribute it and/or modify it under the
 same terms as Perl itself.

data/bin/exiftool

@@ -10,7 +10,7 @@
 use strict;
 require 5.004;
 
-my $version = '12.09';
+my $version = '12.14';
 
 # add our 'lib' directory to the include list BEFORE 'use Image::ExifTool'
 my $exeDir;
@@ -106,6 +106,7 @@ my %csvTags;        # lookup for all found tags with CSV option (lower case keys
 my %database;       # lookup for database information based on file name (in ExifTool Charset)
 my %filterExt;      # lookup for filtered extensions
 my %ignore;         # directory names to ignore
+my $ignoreHidden;   # flag to ignore hidden files
 my %preserveTime;   # preserved timestamps for files
 my %printFmt;       # the contents of the print format file
 my %setTags;        # hash of list references for tags to set from files
@@ -136,12 +137,13 @@ my $countSameWr;    # count files written OK but not changed
 my $critical;       # flag for critical operations (disable CTRL-C)
 my $csv;            # flag for CSV option (set to "CSV", or maybe "JSON" when writing)
 my $csvAdd;         # flag to add CSV information to existing lists
+my $csvDelim;       # delimiter for CSV files
 my $csvSaveCount;   # save counter for last CSV file loaded
 my $deleteOrig;     # 0=restore original files, 1=delete originals, 2=delete w/o asking
 my $disableOutput;  # flag to disable normal output
 my $doSetFileName;  # flag set if FileName may be written
 my $doUnzip;        # flag to extract info from .gz and .bz2 files
-my ($end,$endDir);  # flags to end processing
+my ($end,$endDir,%endDir);  # flags to end processing
 my $escapeC;        # C-style escape
 my $escapeHTML;     # flag to escape printed values for html
 my $evalWarning;    # warning from eval
@@ -239,6 +241,7 @@ my %optArgs = (
     '-c' => 1, '-coordformat' => 1,
     '-charset' => 0, # (optional arg; OK because arg cannot begin with "-")
     '-config' => 1,
+    '-csvdelim' => 1,
     '-d' => 1, '-dateformat' => 1,
     '-D' => 0, # necessary to avoid matching lower-case equivalent
     '-echo' => 1, '-echo1' => 1, '-echo2' => 1, '-echo3' => 1, '-echo4' => 1,
@@ -246,11 +249,12 @@ my %optArgs = (
     '-efile!' => 1, '-efile1!' => 1, '-efile2!' => 1, '-efile3!' => 1, '-efile4!' => 1,
     '-ext' => 1, '--ext' => 1, '-ext+' => 1, '--ext+' => 1,
         '-extension' => 1, '--extension' => 1, '-extension+' => 1, '--extension+' => 1,
-    '-fileorder' => 1,
+    '-fileorder' => 1, '-fileorder0' => 1, '-fileorder1' => 1, '-fileorder2' => 1,
+        '-fileorder3' => 1, '-fileorder4' => 1, '-fileorder5' => 1,
     '-geotag' => 1,
     '-globaltimeshift' => 1,
     '-i' => 1, '-ignore' => 1,
-    '-if' => 1, '-if0' => 1, '-if1' => 1, '-if2' => 1, '-if3' => 1, '-if4' => 1,
+    '-if' => 1, '-if0' => 1, '-if1' => 1, '-if2' => 1, '-if3' => 1, '-if4' => 1, '-if5' => 1,
     '-lang' => 0, # (optional arg; cannot begin with "-")
     '-listitem' => 1,
     '-o' => 1, '-out' => 1,
@@ -366,13 +370,21 @@ if (grep /^-common_args$/i, @ARGV) {
 # loop over sets of command-line arguments separated by "-execute"
 Command: for (;;) {
 
-@echo3 and print STDOUT join("\n", @echo3), "\n";
-@echo4 and print STDERR join("\n", @echo4), "\n";
+if (@echo3) {
+    my $str = join "\n", @echo3, "\n";
+    $str =~ s/\$\{status\}/$rtnVal/ig;
+    print STDOUT $str;
+}
+if (@echo4) {
+    my $str = join "\n", @echo4, "\n";
+    $str =~ s/\$\{status\}/$rtnVal/ig;
+    print STDERR $str;
+}
 
 $rafStdin->Close() if $rafStdin;
 undef $rafStdin;
 
-# save or previous return codes
+# save our previous return codes
 $rtnValPrev = $rtnVal;
 $rtnValApp = $rtnVal if $rtnVal;
 
@@ -387,7 +399,7 @@ if ($binaryStdout) {
 
 # flush console and print "{ready}" message if -stay_open is in effect
 if ($stayOpen >= 2) {
-    if ($quiet) {
+    if ($quiet and not defined $executeID) {
         # flush output if possible
         eval { require IO::Handle } and STDERR->flush(), STDOUT->flush();
     } else {
@@ -421,6 +433,7 @@ undef %countLink;
 undef %created;
 undef %csvTags;
 undef %database;
+undef %endDir;
 undef %filterExt;
 undef %ignore;
 undef %printFmt;
@@ -456,6 +469,7 @@ undef $fileHeader;
 undef $filtered;
 undef $fixLen;
 undef $forcePrint;
+undef $ignoreHidden;
 undef $joinLists;
 undef $langOpt;
 undef $listItem;
@@ -492,6 +506,7 @@ $countGoodCr = 0;
 $countGoodWr = 0;
 $countNewDir = 0;
 $countSameWr = 0;
+$csvDelim = ',';
 $csvSaveCount = 0;
 $fileTrailer = '';
 $filterFlag = 0;
@@ -554,7 +569,7 @@ if (not $preserveTime and $^O eq 'MSWin32') {
 for (;;) {
 
   # execute the command now if no more arguments or -execute is used
-  if (not @ARGV or ($ARGV[0] =~ /^(-|\xe2\x88\x92)execute(\d*)$/i and not $endOfOpts)) {
+  if (not @ARGV or ($ARGV[0] =~ /^(-|\xe2\x88\x92)execute(\d+)?$/i and not $endOfOpts)) {
     if (@ARGV) {
         $executeID = $2;        # save -execute number for "{ready}" response
         $helped = 1;            # don't show help if we used -execute
@@ -813,7 +828,7 @@ for (;;) {
         next;
     }
     /^config$/i and Warn("Ignored -config option (not first on command line)\n"), shift, next;
-    if (/^csv(\+?=.*)?/i) {
+    if (/^csv(\+?=.*)?$/i) {
         my $csvFile = $1;
         # must process on 2nd pass so -f and -charset options are available
         unless ($pass) {
@@ -833,7 +848,7 @@ for (;;) {
             if ($mt->Open(\*CSVFILE, $csvFile)) {
                 binmode CSVFILE;
                 require Image::ExifTool::Import;
-                $msg = Image::ExifTool::Import::ReadCSV(\*CSVFILE, \%database, $forcePrint);
+                $msg = Image::ExifTool::Import::ReadCSV(\*CSVFILE, \%database, $forcePrint, $csvDelim);
                 close(CSVFILE);
             } else {
                 $msg = "Error opening CSV file '${csvFile}'";
@@ -844,6 +859,15 @@ for (;;) {
         $csv = 'CSV';
         next;
     }
+    if (/^csvdelim$/i) {
+        $csvDelim = shift;
+        defined $csvDelim or Error("Expecting argument for -csvDelim option\n"), $badCmd=1, next;
+        $csvDelim =~ /"/ and Error("CSV delimiter can not contain a double quote\n"), $badCmd=1, next;
+        my %unescape = ( 't'=>"\t", 'n'=>"\n", 'r'=>"\r", '\\' => '\\' );
+        $csvDelim =~ s/\\(.)/$unescape{$1}||"\\$1"/sge;
+        $mt->Options(CSVDelim => $csvDelim);
+        next;
+    }
     if (/^d$/ or $a eq 'dateformat') {
         my $fmt = shift;
         $fmt or Error("Expecting date format for -d option\n"), $badCmd=1, next;
@@ -964,6 +988,7 @@ for (;;) {
         my $dir = shift;
         defined $dir or Error("Expecting directory name for -i option\n"), $badCmd=1, next;
         $ignore{$dir} = 1;
+        $dir eq 'HIDDEN' and $ignoreHidden = 1;
         next;
     }
     if (/^if(\d*)$/i) {
@@ -1841,7 +1866,7 @@ if (defined $deleteOrig) {
         printf($o "%5d image files %s\n", $countCopyWr, $overwriteOrig ? 'moved' : 'copied') if $countCopyWr;
         printf($o "%5d files weren't updated due to errors\n", $countBadWr) if $countBadWr;
         printf($o "%5d files weren't created due to errors\n", $countBadCr) if $countBadCr;
-        printf($o "%5d image files read\n", $count) if $tot>1 or ($countDir and not $totWr);
+        printf($o "%5d image files read\n", $count) if ($tot+$countFailed)>1 or ($countDir and not $totWr);
         printf($o "%5d files could not be read\n", $countBad) if $countBad;
         printf($o "%5d output files created\n", scalar(keys %created)) if $textOut;
         printf($o "%5d output files appended\n", scalar(keys %appended)) if %appended;
@@ -1974,6 +1999,7 @@ sub GetImageInfo($$)
                 }
                 last unless $result;
             }
+            undef @foundTags if $fastCondition; # ignore if we didn't get all tags
         }
         unless ($result) {
             $verbose and print $vout "-------- $file (failed condition)$progStr\n";
@@ -2227,7 +2253,9 @@ sub GetImageInfo($$)
         my $noDups = ($json or ($xml and $outFormat > 0));
         my $printConv = $et->Options('PrintConv');
         my $lastGroup = '';
+        my $i = -1;
 TAG:    foreach $tag (@foundTags) {
+            ++$i;   # keep track on index in @foundTags
             my $tagName = GetTagName($tag);
             my ($group, $valList);
             # get the value for this tag
@@ -2345,26 +2373,38 @@ TAG:    foreach $tag (@foundTags) {
             }
             # save information for CSV output
             if ($csv) {
-                my $t = $group ? "$group:$tagName" : $tagName;
-                $t .= '#' if $tag =~ /#/;   # add ValueConv "#" suffix if used
+                my $tn = $tagName;
+                $tn .= '#' if $tag =~ /#/;  # add ValueConv "#" suffix if used
+                my $gt = $group ? "$group:$tn" : $tn;
                 # (tag-name case may be different if some tags don't exist
                 # in a file, so all logic must use lower-case tag names)
-                my $lcTag = lc $t;
+                my $lcTag = lc $gt;
                 # override existing entry only if top priority
                 next if defined $csvInfo{$lcTag} and $tag =~ /\(/;
                 $csvInfo{$lcTag} = $val;
                 if (defined $csvTags{$lcTag}) {
                     # overwrite with actual extracted tag name
                     # (note: can't check "if defined $val" here because -f may be used)
-                    $csvTags{$lcTag} = $t if defined $$info{$tag};
-                } else {
-                    # (don't save unextracted tag name unless -f was used)
-                    $csvTags{$lcTag} = defined($val) ? $t : '';
-                    if (@csvFiles == 1) {
-                        push @csvTags, $lcTag; # save order of tags for first file
-                    } elsif (@csvTags) {
-                        undef @csvTags;
-                    }
+                    $csvTags{$lcTag} = $gt if defined $$info{$tag};
+                    next;
+                }
+                # must check for "Unknown" group (for tags that don't exist)
+                if ($group and defined $csvTags[$i] and $csvTags[$i] =~ /^(.*):$tn$/i) {
+                     next if $group eq 'Unknown';   # nothing more to do if we don't know tag group
+                     if ($1 eq 'unknown') {
+                        # replace unknown entry in CSV tag lookup and list
+                        delete $csvTags{$csvTags[$i]};
+                        $csvTags{$lcTag} = defined($val) ? $gt : '';
+                        $csvTags[$i] = $lcTag;
+                        next;
+                     }
+                }
+                # (don't save unextracted tag name unless -f was used)
+                $csvTags{$lcTag} = defined($val) ? $gt : '';
+                if (@csvFiles == 1) {
+                    push @csvTags, $lcTag; # save order of tags for first file
+                } elsif (@csvTags) {
+                    undef @csvTags;
                 }
                 next;
             }
@@ -2818,6 +2858,10 @@ sub SetImageInfo($$$)
         # determine what our output file name should be
         my $newFileName = $et->GetNewValues('FileName');
         my $newDir = $et->GetNewValues('Directory');
+        if (defined $newFileName and not length $newFileName) {
+            Warn "Warning: New file name is empty - $ infile\n";
+            undef $newFileName;
+        }
         if (defined $testName) {
             my $err;
             $err = "You shouldn't write FileName or Directory with TestFile" if defined $newFileName or defined $newDir;
@@ -2896,14 +2940,19 @@ sub SetImageInfo($$$)
             }
             # quickly rename file and/or set file date if this is all we are doing
             if ($numSet == $numPseudo) {
-                my $r1 = $et->SetFileModifyDate($file,undef,'FileCreateDate');
-                my $r2 = $et->SetFileModifyDate($file);
-                my $r3 = $et->SetSystemTags($file);
-                my $r4 = 0;
-                $r4 = $et->SetFileName($file, $outfile) if defined $outfile;
-                if ($r1 > 0 or $r2 > 0 or $r3 > 0 or $r4 > 0) {
+                my ($r0, $r1, $r2, $r3) = (0, 0, 0, 0);
+                if (defined $outfile) {
+                    $r0 = $et->SetFileName($file, $outfile);
+                    $file = $$et{NewName} if $r0 > 0;   # continue with new name if changed
+                }
+                unless ($r0 < 0) {
+                    $r1 = $et->SetFileModifyDate($file,undef,'FileCreateDate');
+                    $r2 = $et->SetFileModifyDate($file);
+                    $r3 = $et->SetSystemTags($file);
+                }
+                if ($r0 > 0 or $r1 > 0 or $r2 > 0 or $r3 > 0) {
                     ++$countGoodWr;
-                } elsif ($r1 < 0 or $r2 < 0 or $r3 < 0 or $r4 < 0) {
+                } elsif ($r0 < 0 or $r1 < 0 or $r2 < 0 or $r3 < 0) {
                     EFile($infile);
                     ++$countBadWr;
                     return 0;
@@ -2912,8 +2961,7 @@ sub SetImageInfo($$$)
                     ++$countSameWr;
                 }
                 if (defined $hardLink or defined $symLink or defined $testName) {
-                    my $src = (defined $outfile and $r4 > 0) ? $outfile : $file;
-                    DoHardLink($et, $src, $hardLink, $symLink, $testName);
+                    DoHardLink($et, $file, $hardLink, $symLink, $testName);
                 }
                 return 1;
             }
@@ -3290,7 +3338,7 @@ sub FormatCSV($)
     # the -b option is used to encode as Base64.  It is unclear whether or not this
     # is valid CSV, but some readers may not like it.  (If this becomes a problem,
     # in the future values may need to be truncated at the first NULL character.)
-    $val = qq{"$val"} if $val =~ s/"/""/g or $val =~ /(^\s+|\s+$)/ or $val =~ /[,\n\r]/;
+    $val = qq{"$val"} if $val =~ s/"/""/g or $val =~ /(^\s+|\s+$)/ or $val =~ /[\n\r]|\Q$csvDelim/;
     return $val;
 }
 
@@ -3303,9 +3351,9 @@ sub PrintCSV()
     @csvTags or @csvTags = sort keys %csvTags;
     # make a list of tags actually found
     foreach $lcTag (@csvTags) {
-        push @tags, $csvTags{$lcTag} if $csvTags{$lcTag};
+        push @tags, FormatCSV($csvTags{$lcTag}) if $csvTags{$lcTag};
     }
-    print join(',', 'SourceFile', @tags), "\n";
+    print join($csvDelim, 'SourceFile', @tags), "\n";
     my $empty = defined($forcePrint) ? $forcePrint : '';
     foreach $file (@csvFiles) {
         my @vals = (FormatCSV($file)); # start with full file name
@@ -3316,7 +3364,7 @@ sub PrintCSV()
             defined $val or push(@vals,$empty), next;
             push @vals, FormatCSV($val);
         }
-        print join(',', @vals), "\n";
+        print join($csvDelim, @vals), "\n";
     }
 }
 
@@ -3552,8 +3600,18 @@ sub ProcessFiles($;$)
             if ($list) {
                 push(@$list, $file);
             } else {
+                if (%endDir) {
+                    my ($d, $f) = Image::ExifTool::SplitFileName($file);
+                    next if $endDir{$d};
+                }
                 GetImageInfo($et, $file);
                 $end and Warn("End called - $file\n");
+                if ($endDir) {
+                    Warn("EndDir called - $file\n");
+                    my ($d, $f) = Image::ExifTool::SplitFileName($file);
+                    $endDir{$d} = 1;
+                    undef $endDir;
+                }
             }
         }
         $et->Options(CharsetFileName => $enc) if $utf8FileName{$file};
@@ -3639,6 +3697,7 @@ sub ScanDir($$;$)
             next;
         }
         next if $endThisDir;
+        next if $ignoreHidden and $file =~ /^\./;   # ignore hidden files if specified
         # apply rules from -ext options
         my $accepted;
         if ($filterFlag) {
@@ -3683,8 +3742,9 @@ sub ScanDir($$;$)
             if ($endDir) {
                 $path =~ s(/$)();
                 Warn("EndDir called - $path\n");
-                undef $endDir;
+                $endDir{$path} = 1;
                 $endThisDir = 1;
+                undef $endDir;
             }
         }
     }
@@ -4478,6 +4538,7 @@ L<Input-output text formatting|/Input-output text formatting>
   -c FMT      (-coordFormat)       Set format for GPS coordinates
   -charset [[TYPE=]CHARSET]        Specify encoding for special characters
   -csv[[+]=CSVFILE]                Export/import tags in CSV format
+  -csvDelim STR                    Set delimiter for CSV file
   -d FMT      (-dateFormat)        Set format for date/time values
   -D          (-decimal)           Show tag ID numbers in decimal
   -E,-ex,-ec  (-escape(HTML|XML|C))Escape tag values for HTML, XML or C
@@ -4997,8 +5058,11 @@ tag names (with optional group names) for each column of the file, and
 values must be separated by commas.  A special "SourceFile" column specifies
 the files associated with each row of information (and a SourceFile of "*"
 may be used to define default tags to be imported for all files which are
-combined with any tags specified for the specific SourceFile processed).
-The following examples demonstrate basic use of this option:
+combined with any tags specified for the specific SourceFile processed). The
+B<-csvDelim> option may be used to change the input/output field delimiter
+if something other than a comma is required.
+
+The following examples demonstrate basic use of the B<-csv> option:
 
     # generate CSV file with common tags from all images in a directory
     exiftool -common -csv dir > out.csv
@@ -5045,6 +5109,13 @@ option.  When processing a large number of files, it is recommended to
 either use the JSON (B<-j>) or XML (B<-X>) output format, or use B<-p> to
 generate a fixed-column CSV file instead of using the B<-csv> option.
 
+=item B<-csvDelim> I<STR>
+
+Set the delimiter for separating CSV entries for CSV file input/output via
+the B<-csv> option.  I<STR> may contain "\t", "\n", "\r" and "\\" to
+represent TAB, LF, CR and '\' respectively.  A double quote is not allowed
+in the delimiter.  Default is ','.
+
 =item B<-d> I<FMT> (B<-dateFormat>)
 
 Set the format for date/time tag values.  The I<FMT> string may contain
@@ -5296,7 +5367,7 @@ with this command:
 
 produces output like this:
 
-    -- Generated by ExifTool 12.09 --
+    -- Generated by ExifTool 12.14 --
     File: a.jpg - 2003:10:31 15:44:19
     (f/5.6, 1/60s, ISO 100)
     File: b.jpg - 2006:05:23 11:57:38
@@ -5704,7 +5775,8 @@ extracting any EXIF MakerNote information.  B<-fast3> avoids extracting
 metadata from the file, and returns only pseudo System tags, but still reads
 the file header to obtain an educated guess at FileType.  B<-fast4> doesn't
 even read the file header, and returns only System tags and a FileType based
-on the file extension.  Has no effect when writing.
+on the file extension.  B<-fast5> also disables generation of the Composite
+tags (like B<-e>).  Has no effect when writing.
 
 Note that a separate B<-fast> setting may be used for evaluation of a B<-if>
 condition, or when ordering files with the B<-fileOrder> option.  See the
@@ -5729,7 +5801,9 @@ involves an additional processing pass of each file, but this impact may be
 reduced by specifying a I<NUM> for the B<-fast> level used during the
 metadata-extraction phase.  For example, B<-fileOrder4> may be used if
 I<TAG> is a pseudo System tag.  If multiple B<-fileOrder> options are used,
-the extraction is done at the lowest B<-fast> level.
+the extraction is done at the lowest B<-fast> level.  Note that files are
+sorted across directory boundaries if multiple input directories are
+specified.
 
 =item B<-i> I<DIR> (B<-ignore>)
 
@@ -5738,7 +5812,9 @@ name, or a full path.  If a full path is specified, it must match the
 Directory tag exactly to be ignored.  Use multiple B<-i> options to ignore
 more than one directory name.  A special I<DIR> value of C<SYMLINKS> (case
 sensitive) may be specified to ignore symbolic links when the B<-r> option
-is used.
+is used.  As well, a value of C<HIDDEN> (case sensitive) may be used to
+ignore files with names that start with a "." (ie. hidden files on Unix
+systems) when scanning a directory.
 
 =item B<-if>[I<NUM>] I<EXPR>
 
@@ -5770,7 +5846,7 @@ Adding I<NUM> to the B<-if> option causes a separate processing pass to be
 executed for evaluating I<EXPR> at a B<-fast> level given by I<NUM> (see the
 B<-fast> option documentation for details).  Without I<NUM>, only one
 processing pass is done at the level specified by the B<-fast> option.  For
-example, using B<-if4> is possible if I<EXPR> uses only pseudo System tags,
+example, using B<-if5> is possible if I<EXPR> uses only pseudo System tags,
 and may significantly speed processing if enough files fail the condition.
 
 The expression has access to the current ExifTool object through C<$self>,
@@ -5779,7 +5855,8 @@ of the file processing.  Both functions have a return value of 1.  Case is
 significant for function names.
 
     End()    - end processing after this file
-    EndDir() - end processing of files in this directory
+    EndDir() - end processing of files in this directory (not
+               compatible with the B<-fileOrder> option)
 
 Notes:
 
@@ -6241,7 +6318,9 @@ about the configuration file syntax.
 Echo I<TEXT> to stdout (B<-echo> or B<-echo1>) or stderr (B<-echo2>).  Text
 is output as the command line is parsed, before the processing of any input
 files.  I<NUM> may also be 3 or 4 to output text (to stdout or stderr
-respectively) after processing is complete.
+respectively) after processing is complete.  For B<-echo3> and B<-echo4>,
+"${status}" may be used in the I<TEXT> string to represent the numerical
+exit status of the command (see L</EXIT STATUS>).
 
 =item B<-efile>[I<NUM>][!] I<ERRFILE>
 
@@ -6262,7 +6341,8 @@ commands were executed as separate command lines (with the exception of the
 B<-config> and B<-use> options which remain in effect for subsequent
 commands).  Allows multiple commands to be executed from a single command
 line.  I<NUM> is an optional number that is echoed in the "{ready}" message
-when using the B<-stay_open> feature.
+when using the B<-stay_open> feature.  If a I<NUM> is specified, the B<-q>
+option no longer suppresses the output "{readyNUM}" message.
 
 =item B<-srcfile> I<FMT>
 
@@ -6293,10 +6373,10 @@ name and directory.
 
 =item B<-stay_open> I<FLAG>
 
-If I<FLAG> is C<1> or C<True>, causes exiftool keep reading from the B<-@>
-I<ARGFILE> even after reaching the end of file.  This feature allows calling
-applications to pre-load exiftool, thus avoiding the overhead of loading
-exiftool for each command.  The procedure is as follows:
+If I<FLAG> is C<1> or C<True> (case insensitive), causes exiftool keep
+reading from the B<-@> I<ARGFILE> even after reaching the end of file.  This
+feature allows calling applications to pre-load exiftool, thus avoiding the
+overhead of loading exiftool for each command.  The procedure is as follows:
 
 1) Execute C<exiftool -stay_open True -@ I<ARGFILE>>, where I<ARGFILE> is the
 name of an existing (possibly empty) argument file or C<-> to pipe arguments
@@ -6312,12 +6392,16 @@ received up to this point, send a "{ready}" message to stdout when done
 (unless the B<-q> or B<-T> option is used), and continue trying to read
 arguments for the next command from I<ARGFILE>.  To aid in command/response
 synchronization, any number appended to the C<-execute> option is echoed in
-the "{ready}" message.  For example, C<-execute613> results in "{ready613}".
+the "{ready}" message.  For example, C<-execute613> results in "{ready613}". 
+When this number is added, B<-q> no longer suppresses the "{ready}" message.
+(Also, see the B<-echo3> and B<-echo4> options for additional ways to pass
+signals back to your application.)
 
 4) Repeat steps 2 and 3 for each command.
 
-5) Write C<-stay_open\nFalse\n> to I<ARGFILE> when done.  This will cause
-exiftool to process any remaining command-line arguments then exit normally.
+5) Write C<-stay_open\nFalse\n> (or C<-stay_open\n0\n>) to I<ARGFILE> when
+done.  This will cause exiftool to process any remaining command-line
+arguments then exit normally.
 
 The input I<ARGFILE> may be changed at any time before step 5 above by
 writing the following lines to the currently open I<ARGFILE>:
@@ -6327,7 +6411,7 @@ writing the following lines to the currently open I<ARGFILE>:
     -@
     NEWARGFILE
 
-This causes I<ARGFILE> to be closed, and I<NEWARGFILE> to be kept open.
+This causes I<ARGFILE> to be closed, and I<NEWARGFILE> to be kept open. 
 (Without the B<-stay_open> here, exiftool would have returned to reading
 arguments from I<ARGFILE> after reaching the end of I<NEWARGFILE>.)
 
@@ -6345,10 +6429,11 @@ an interface to the API UserParam option (see the
 L<Image::ExifTool Options|Image::ExifTool/Options> documentation), and
 provides a method to access user-defined parameters in arguments to the
 B<-if> and B<-p> options as if they were any other tag.  Appending a hash
-tag (C<#>) to I<PARAM> also causes the parameter to be extracted as a normal
-tag (in the UserParam group).  Similar to the B<-api> option, the parameter
-value is set to 1 if I<=VAL> is omitted, undef if just I<VAL> is omitted
-with C<=>, or an empty string if I<VAL> is omitted with C<^=>.
+tag (C<#>) to I<PARAM> (eg. C<-userParam MyTag#=yes>) also causes the
+parameter to be extracted as a normal tag in the UserParam group.  Similar
+to the B<-api> option, the parameter value is set to 1 if I<=VAL> is
+omitted, undef if just I<VAL> is omitted with C<=>, or an empty string if
+I<VAL> is omitted with C<^=>.
 
     exiftool -p '$test from $filename' -userparam test=Hello FILE
 
@@ -6418,9 +6503,9 @@ For example, the following command may be used to remove duplicate Keywords:
 The B<-sep> option is necessary to split the string back into individual
 list items when writing to a list-type tag.
 
-An optional flag argument may be set to 1 to cause C<NoDups> to return undef
-if no duplicates existed, thus preventing the file from being rewritten
-unnecessarily:
+An optional flag argument may be set to 1 to cause C<NoDups> to set C<$_> to
+undef if no duplicates existed, thus preventing the file from being
+rewritten unnecessarily:
 
     exiftool -sep '##' '-keywords<${keywords;NoDups(1)}' a.jpg
 
@@ -7058,6 +7143,13 @@ flexibility of ExifTool.)
 
 =back
 
+=head1 INTERRUPTING EXIFTOOL
+
+Interrupting exiftool with a CTRL-C or SIGINT will not result in partially
+written files or temporary files remaining on the hard disk.  The exiftool
+application traps SIGINT and defers it until the end of critical processes
+if necessary, then does a proper cleanup before exiting.
+
 =head1 EXIT STATUS
 
 The exiftool application exits with a status of 0 on success, or 1 if an
@@ -7066,7 +7158,7 @@ the commands if B<-execute> was used).
 
 =head1 AUTHOR
 
-Copyright 2003-2020, Phil Harvey
+Copyright 2003-2021, Phil Harvey
 
 This is free software; you can redistribute it and/or modify it under the
 same terms as Perl itself.