Installer apache-opennlp avec Homebrew, apt, zypper

Liens du projet

• Page d'accueil
• Dépôt
• Docs amont
• Page du gestionnaire de paquets

Commandes et alias

• opennlp

Historique du projet

OpenNLP existed before Apache incubation, with current Apache pages still pointing to archived SourceForge releases and model resources. The project entered the Apache Incubator in November 2010 and graduated as an Apache top-level project on 2012-02-15.

The incubation record describes the project goal as creating a mature toolkit for common NLP tasks and providing pre-built models for multiple languages. The Apache-era project continued to ship both a Java API and a common opennlp command-line entry point.

Historique d'adoption

OpenNLP's adoption is strongest in Java and JVM text-processing pipelines where developers want a conventional library and CLI for classic NLP rather than a server or hosted model. Its old SourceForge model archive and Apache manual both show the long-running package culture around downloading binary models, running trainers, and wiring the toolkit into UIMA or custom applications.

As newer deep-learning NLP stacks arrived, OpenNLP remained relevant for lightweight, inspectable, trainable components and for organizations with Java pipelines that need sentence splitting, tokenization, tagging, NER, and parser tools without adopting a Python-heavy stack.

Modes d'utilisation

The Homebrew package exposes the opennlp command. Apache's manual says the binary distribution includes an opennlp script for Linux-compatible systems and opennlp.bat for Windows, with OPENNLP_HOME and PATH used to make the tools convenient.

Typical CLI usage invokes a tool name plus a model, reading from stdin and writing to stdout, or runs trainer commands with model output, language, training data, encoding, and task-specific options. That makes the package friendly for shell pipelines and repeatable model-training experiments.

Pourquoi les passionnés de paquets s'y intéressent

OpenNLP is appealing to package nerds because it is a Java NLP library with a real Unix-style CLI surface. You can install it, fetch or provide models, pipe text through tokenizers or taggers, and train task-specific models without building an application first.

It also carries a visible history of pre-Apache open-source distribution: SourceForge-era releases and models, Apache incubation, Maven artifacts, and a CLI that still exposes the toolkit as a collection of small named NLP tools.

Chronologie

• 2010: OpenNLP enters the Apache Incubator.
• 2011: Apache OpenNLP 1.5.2-incubating is released.
• 2012: OpenNLP graduates as an Apache top-level project.
• 2026: Apache documentation lists current 2.x and 3.x milestone manuals and Javadocs.

Related projects

• Apache UIMA is an adjacent Java text-analysis framework with documented OpenNLP integration.
• SourceForge-era OpenNLP models, Maven Central artifacts, ONNX Runtime support, and Java NLP pipelines are common adjacent pieces.

Sources

• https://incubator.apache.org/projects/opennlp.html
• https://opennlp.apache.org/docs/2.5.4/manual/opennlp.html
• https://opennlp.apache.org/download.html
• https://opennlp.apache.org/models.html
• https://opennlp.apache.org/news/news-2010-11-23.html