This is a simple tool to split a German compound noun into separate sub-nouns using a wordlist.

If no wordlist is specified via the -d option, data/german-common-nouns-extended.txt must be present relative to the execution directory.

To enable runtime measuring, set environment variable PROFILE to true. Profiling information is printed to stderr, so it won't interfere with piping the output to other processes.To replace umlauts with regular characters, set the environment variable PURGE_UMLAUT to true. This is necessary for some wordlists.

Example Usage:

java -jar bin/composite-split.jar Donaudampfschifffahrtselektrizitätenhauptbetriebswerkbauunterbeamtengesellschaft

java -jar -d path/to/my/wordlist.txt Donaudampfschifffahrtselektrizitätenhauptbetriebswerkbauunterbeamtengesellschaft

Additionally, there is a test.sh script that runs all compound words specified in data/compounds.txt.

• Extend wordlist with -ung form of common verbs
• Support irregular pluralization
• Favor shorter words if this allows finding a following word:

The code in this repository is covered by the LICENSE file, except for the files in data. Those files have been taken from various sources under the terms of academic fair use. If you are the copyright holder and object the use of those files in this context, please get in touch so they can be taken down.