Dockerfile for LanguageTool

This repository contains a Dockerfile to create a Docker image for LanguageTool.

LanguageTool is an Open Source proofreading software for English, French, German, Polish, Russian, and more than 20 other languages. It finds many errors that a simple spell checker cannot detect.

Setup

Setup using Docker Hub

docker pull erikvl87/languagetool 
 docker run --rm -p 8010:8010 erikvl87/languagetool

This will pull the latest tag from Docker Hub. Optionally, specify a tag to pin onto a fixed version. These versions are derived from the official LanguageTool releases. Updates to the Dockerfile for already published versions are released with a -dockerupdate-{X} postfix in the tag (where {X} is an incremental number).

Setup using the Dockerfile

This approach could be used when you plan to make changes to the Dockerfile.

git clone https://github.com/Erikvl87/docker-languagetool.git --config core.autocrlf=input 
 docker build -t languagetool . 
 docker run --rm -it -p 8010:8010 languagetool

Configuration

Java heap size

LanguageTool will be started with a minimal heap size (-Xms) of 256m and a maximum (-Xmx) of 512m. You can overwrite these defaults by setting the environment variables JavaXms and Java_Xmx. An example startup configuration:

docker run --rm -it -p 8010:8010 -e JavaXms=512m -e JavaXmx=2g erikvl87/languagetool

LanguageTool HTTPServerConfig

You are able to use the HTTPServerConfig configuration options by prefixing the fields with langtool and setting them as environment variables. An example startup configuration:

docker run --rm -it -p 8010:8010 -e langtoolpipelinePrewarming=true -e JavaXms=1g -e Java_Xmx=2g erikvl87/languagetool

FastText support

FastText is preinstalled in the image and the lid.176.bin model is downloaded automatically. Because the image wires these values internally, providing langtoolfasttextModel or langtoolfasttextBinary environment variables causes the container to exit with an error.

Overwrite logging configuration

To overwrite the default LanguageTool logback.xml logging configuration, create a new logback.xml file and mount it into the container. For example, create the following logback.xml file:

<?xml version="1.0" encoding="UTF-8"?> 
 <configuration> 
   <logger name="org.languagetool" level="ERROR"/> 
 </configuration>

An example startup configuration:

docker run --rm -it -p 8010:8010 -v /home/john/logback.xml:/LanguageTool/logback.xml erikvl87/languagetool

Using n-gram datasets

LanguageTool can make use of large n-gram data sets to detect errors with words that are often confused, like their and there.

Source: https://dev.languagetool.org/finding-errors-using-n-gram-data Download the n-gram dataset(s) onto your local machine and unzip them into a local ngrams directory:

home/ 
 ├─ john/ 
 │  ├─ ngrams/ 
 │  │  ├─ en/ 
 │  │  │  ├─ 1grams/ 
 │  │  │  ├─ 2grams/ 
 │  │  │  ├─ 3grams/ 
 │  │  ├─ nl/ 
 │  │  │  ├─ 1grams/ 
 │  │  │  ├─ 2grams/ 
 │  │  │  ├─ 3grams/

Mount the local ngrams directory to the /ngrams directory in the Docker container using the -v configuration and set the languageModel configuration to the /ngrams folder. An example startup configuration:

docker run --rm -it -p 8010:8010 -e langtool_languageModel=/ngrams -v /home/john/ngrams:/ngrams:ro erikvl87/languagetool

Improving the spell checker

You can improve the spell checker without touching the dictionary. For single words (no spaces), you can add your words to one of these files:

* spelling.txt: words that the spell checker will ignore and use to generate corrections if someone types a similar word

* ignore.txt: words that the spell checker will ignore but not use to generate corrections

* prohibited.txt: words that should be considered incorrect even though the spell checker would accept them

Source: https://dev.languagetool.org/hunspell-support The following Dockerfile contains an example on how to add words to spelling.txt. It assumes you have your own list of words in enspellingadditions.txt next to the Dockerfile.

FROM erikvl87/languagetool 
  
 
Improving the spell checker 
 
http://wiki.languagetool.org/hunspell-support 
 USER root 
 COPY enspellingadditions.txt enspellingadditions.txt 
 RUN  (echo; cat enspellingadditions.txt) >> org/languagetool/resource/en/hunspell/spelling.txt 
 USER languagetool

You can build & run the custom Dockerfile with the following two commands:

docker build -t languagetool-custom . 
 docker run --rm -it -p 8010:8010 languagetool-custom

You can add words to other languages by changing the en language tag in the target path. Note that for some languages, e.g. for nl the spelling.txt file is not in the hunspell folder: org/languagetool/resource/nl/spelling/spelling.txt.

Docker Compose

This image can also be used with Docker Compose. An example docker-compose.yml is located at the root of this project.

Usage

By default this image is configured to listen on port 8010 which deviates from the default port of LanguageTool 8081. An example cURL request:

curl --data "language=en-US&text=a simple test" http://localhost:8010/v2/check

Please refer to the official LanguageTool documentation for further usage instructions.

Known issues & workarounds

If you experience problems when connecting local server to the official Firefox extension, see cors-workaround.