Interactive mode (REPL)

Built-in interactive mode

You can type or paste sentences or paragraphs into CoreNLP interactively and see how it analyzes them. This gives a kind of read-eval-print loop (REPL).

You can use interactive mode with either StanfordCoreNLP or the combination of running StanfordCoreNLPServer and StanfordCoreNLPClient. You can specify whatever annotators and other properties that you want.

If you do not specify any flag that directs CoreNLP to process text from a particular file, then after the pipeline is loaded, you will be placed into an interactive loop. (That is, if you don’t specify either -file or -filelist.)

You exit the REPL by typing:

q RETURN

Typing CTRL-C also works!

Java 9 REPL

If you are trying out the REPL for Java 9, you might want to try out using the CoreNLP simple API with it. It works quite well.