Watch a directory and process new text files with Stanford CoreNLP
CoreNLP-Watcher is a small Java application that watches a given directory and automatically processses any new text files that appear there, using the Stanford CoreNLP tools. This is may be useful if you want to quckly process new text files as they appear in real time (for example, when running a web application) without the overhead of loading the CoreNLP models each time you process a new file, and/or if you want to utilize CoreNLP from an application that is not written in Java, using a simple file-based protocol.
You need to have the Stanford CoreNLP version 3.5.1 (download link) downloaded and unpacked somewhere on your computer.
Note: CoreNLP version 3.6.0 introduced a built-in API server, which largely renders CoreNLP-Watcher obsolete.
If do not have Stanford CoreNLP installed already, download and unzip the package somewhere on your computer.
Edit the settings file with the full path to the CoreNLP files, as
well as the full path to the directory to be watched.
To compile the CoreNLPWatcher Java code, use the provided
Makefile. Note that this requries that the path to CoreNLP be
specified in an environment variable, which can be done by sourcing
the settings file.
source settings
make
Launch CoreNLPWatchr by executing the corenlpw script. You will need
to wait a minute or so for the CoreNLP models to load into memory, and
then the watcher will automatically process any new files that appear
in the watch direct.
./corenlpw &
For example, if the watcher is running and watching the
/tmp/corenlpw directory, you can test it like so:
echo "This is a test." > /tmp/corenlpw/test.txt
Which should result in a file watch/test.txt.xml containing the
CoreNLP annotation output.
As it incorporates the Stanford CoreNLP code, this software is licensed under the GNU General Public License.
By Terrence Szymanski. www.affrication.org