Using metadata-extractor from JRuby

# this file's in JRuby, which is just Ruby that can access Java. it's based on
# code samples here: http://www.drewnoakes.com/code/exif/sampleUsage.html

# you can download/install JRuby to run this code, or just use it as a guide to translate into
# Java. it's mostly just copied from the sample code; I just used JRuby because I figured it'd
# be quicker than Java as I only needed to demo libraries and make sure they worked nicely.

# the stable release handles JPEGs, and has a great track record (used in Mathematica for example).
# http://www.drewnoakes.com/code/exif/releases/metadata-extractor-2.3.1.jar

require 'java'

# file
poseidon = java.io.File.new("metadata/Poseidon0466.JPG")

# extract metadata container object from file
meta_poseidon = com.drew.imaging.ImageMetadataReader.read_metadata(poseidon)
exif_reader_with_poseidon = com.drew.metadata.exif.ExifReader.new(poseidon)
extracted_metadata = exif_reader_with_poseidon.extract(meta_poseidon)

# this is weird to me but basically the metadata's stored in directories of tags, so to
# get the actual information, you need to get directories and then pull out the tags.
exif_dir = extracted_metadata.getDirectory(com.drew.metadata.exif.ExifDirectory.java_class)
iptc_dir = extracted_metadata.getDirectory(com.drew.metadata.iptc.IptcDirectory.java_class)

# get some specific tags
make = exif_dir.get_string(com.drew.metadata.exif.ExifDirectory::TAG_MAKE)
model = exif_dir.get_string(com.drew.metadata.exif.ExifDirectory::TAG_MODEL)
caption = iptc_dir.get_string(com.drew.metadata.iptc.IptcDirectory::TAG_CAPTION)

# print to STDOUT

puts "\n\n\n\nmake, model, caption:\n\n\n\n"
puts make
puts model
puts caption

# loop over all tags - I'm not sure this Java-esque approach to iterating is really necessary here,
# but since it's going to be translated into Java anyway, it probably does no harm

def print_all_tags(metadata)
  java_style_directory_iterator = metadata.get_directory_iterator
  while java_style_directory_iterator.has_next
    tag_directory = java_style_directory_iterator.next
    java_style_tag_iterator = tag_directory.get_tag_iterator
    while java_style_tag_iterator.has_next
      tag = java_style_tag_iterator.next
      puts tag
    end
  end
end
puts "\n\n\n\nlooping over jpeg\n\n\n\n"
print_all_tags(meta_poseidon)

# abort "skipping TIFFs for now"

# the TIFF functionality is in beta, but it appears to be a GMail kind of beta. the library appears to
# work perfectly for these sample files.
# http://www.drewnoakes.com/code/exif/releases/metadata-extractor-2.4.0-beta-1.jar

tiff_filename = "metadata/100+Girls+redux+Times+test.tif"

%w{metadata/100+Girls+redux+Times+test.tif
   metadata/Army+of+Darkness+bad+guys.tif
   metadata/ER+Anthony+Edwards+hands.tif}.each do |tiff_filename|
  puts "\n\n\n\nlooping over #{tiff_filename}\n\n\n\n"
  print_all_tags(com.drew.imaging.ImageMetadataReader.read_metadata(java.io.File.new(tiff_filename)))
end

# the looping code is much easier to read, and the tags present in the metadata for our sample files varies
# somewhat unpredictably, so I'd say the looping approach is the best one. one caveat: TIFF metadata includes
# a lot of numeric noise.