NLP utils based on protobuf and grpc

Use docker to launch server

• docker/run.sh defines a few variables, including Stanford CoreNLP version. Modify it as needed.

Run server

• Before running docker/run.sh, make sure you change the three ports to your own ones, and also use a your own container name specified by --name
• Run bash docker/run.sh. This should bring up BLLIP server, nlputils server and visual web interface in a few minutes.
• You can bring up multiple servers using docker/run.sh, as long as you specify different ports each time.
• The server will use IP and ports of the host machine.
• docker logs CONTAINER_NAME to view the terminal outputs, useful for checking if the server is ready
• You can also use the visual web interface to check if the server is ready. Input a simple sentence and see if you get the parsing results.

Stop server

• To stop the server: docker rm -f CONTAINER_NAME

Preparation

• Create a folder under /data/Applications/user_name/ (Disk I/O under this folder is much faster than home directory)
• cd /data/Applications/user_name/ and follow the steps below

Check out the repository

• git clone https://github.com/leebird/nlputils/

Initialization

• Only bash is supported. If you default shell is not bash, run bash before running the following steps.
• First create a python virtual enviroment: virtualenv --python=python2 env. Note that the virtual enviroment can be placed anywhere, not necessarily under nlputils. If you just want to use nlputils as a module, maybe you want to create the virtual environment in the parent folder of nlputils.
• Activate the virtual enviroment: . env/bin/activate
• bash ./scripts/clear_init.sh
• Set up environment path variables: . ./scripts/export_path.sh
• Run the initialization script: bash ./scripts/init.sh
• Compile document.proto to generate python codes: bash ./scripts/compile_document_proto.sh

Example

• Only bash is supported. If you default shell is not bash, run bash before running the following steps.
• Activate the virtual enviroment: . env/bin/activate
• Set up environment path variables: . ./scripts/export_path.sh. Note that these variables can be written into .bashrc so that they are set up when the terminal starts.
• Run python test/example.py, it should print a list of structured information.

Note

• Tested under bash, but not tcsh.