Recently I have been tasked with automatically compiling, tagging and indexing images added to an application via FTP. The images had been tagged using Adobe Bridge or Photoshop, anyone with knowledge of Adobe products will be familiar with tagging images. For those who aren’t, you can add descriptions, titles, ratings, keywords and copyrights (to name a few) to the core of an image using Bridge or Photoshop. My challenge was extracting this information at the other end. As the user would be uploading 1000s of images, there was absolutely no way they would want to tag each image after waiting for it to upload via a web form. After a few cron scripts dumping errors and FTP permission issues I managed to configure a system that would listen in a drop folder on one of the servers and then chmod and move the file to a web friendly directory for processing. This is where the interesting stuff begins.

More >