Currently, analysis does not distinguish between a character name and a character ID for lines. These should be distinguished.