OCR random images using Google Cloud Vision and put the results in the Finder metadata for macOS Spotlight search

ocrimg.rb

# Ruby script to OCR random JPEG and PNG files using Google Cloud Vision, 
# then add the results to the macOS Finder metadata for Spotlight search purposes.

# Before you can use this you need to set up billing for Google Cloud Vision, and
# download the public/private key credentials as a JSON file. See:
# https://cloud.google.com/vision/docs/setup

ENV["GOOGLE_APPLICATION_CREDENTIALS"] = "path-to-google-cloud-vision-credentials-file.json"

require "shellwords"
require "google/cloud/vision"
require "cfpropertylist"
require "find"
require "pp"

PROJECT_ID = "ocr-random-images"

def ocrfile(image_path)
	vision = Google::Cloud::Vision.new project: PROJECT_ID
	image  = vision.image image_path
	document = image.document
	return document.text.gsub(/[\r\n]+/, " ").gsub(/  +/, " ")
end

def plistize(str)
	plist = CFPropertyList::List.new
	plist.value = CFPropertyList.guess(str)
	return plist.to_str
end

# This doesn't seem to work for some unknown reason?
def setcomment(fspc, txt)
	comment = txt.gsub(/[\r\n]/, ' ').gsub(/\s\s*/, " ")
	plist = plistize(comment).unpack("H*")[0]
	system("xattr", "-wx", "com.apple.metadata:kMDItemFinderComment", plist, fspc)
end

def findersetcomment(fspc, txt)
	deq = txt.gsub(/[\r\n]+/, ' ').gsub(/\s\s*/, " ").gsub(/["']/, "")
	system("osascript",
				 "-e",
				 "tell app \"Finder\" to set comment of (POSIX file \"#{fspc}\" as alias) to \"#{deq}\"")
end

def ocrimg(fspc)
	txt = ocrfile(fspc)
	findersetcomment(fspc, txt)
	puts "#{fspc}: #{txt}"
end

def ocrimgs(folder)
	Find.find(folder) do |path|
		if path.match(/\.(jpg|jpeg|png)$/i)
			ocrimg(path)
		end
	end
end

# Now call on a folder full of images
ocrimgs("/Users/meta/Pictures/Cartoons/Scenes From A Multiverse")