Text Nodes are used for retrieving metadata from Versions and for several types of text manipulation such as replacing text, trimming text lengths, capitalising words etc.
A token is a metadata name enclosed in curly brackets, for example '{Aperture}'. When the workflow is run, this will be replaced by the actual value, for example '5.6'. In the Text Node you can access all these tokens via the menus and the Insert button, but you can also type the tokens in directly yourself.
Get Text is used to retrieve metadata from the current Version and to add in your own text. Like the Decision Node, you can choose from EXIF, IPTC, Keywords and Other metadata.
Maximum Length allows you to limit the length of text from previous Nodes in the workflow to the specified number of characters. Spaces will counted as characters.
UPPER/lower/Caps will take the text from previous Nodes in the workflow and turn it into all upper case or all lower case, or set the first character of each word to upper case.