classifymedia

These steps comprise the pipeline used to get files indexed as well as to generate thumbnails for them. Each step has a Start method which is called either by main or by the step it depends on (the one earlier in the pipeline). In addition, each step waits for its own queue to be empty AND for the preceding steps to be done.

In other words, scanner starts checkindex and then waits for checkindex to complete.

The media (add to ElasticSearch) pipeline is:

scanner:

• Scans the file systems for files to examine
• Passes the files to checkindex

checkindex:

• Checks ElasticSearch to determine if the file is already indexed or has changed since it was indexed.
• Passes the files to getexif
  • If it's an update, it passes to generatethumbnail
  • If it's in the index and not updated, it passes to checkthumbnail
  • If it's NOT in the index, it'll be checked after the preparemedia step

getexif:

• Invokes exiftool once for a directory, getting all exif info for each file.
• Passes the data to preparemedia

preparemedia:

• Parses the exif data, creating the media document
• Passes the data to resolveplacename
• Passes the data to checkthumbnail

resolveplacename:

• If the media has a location, do a reverse geocode to get the placename
• Passes the data to indexmedia

indexmedia:

• Adds/updates the media in the index, calling ElasticSearch
• < nothing else >

The thumbnail pipeline is:

checkthumbnail:

• Checks if the thumbnail exists
• Passes to generatethumbnail if not

generatethumbnail:

• Generates the thumbnail
• < nothing else >