Overview

The cas_visualizerlibrary can be used to transform a Common Analysis System (CAS) data structure into an annotated HTML string.

The library visualizes annotations with an underlined (default) or highlighted style.

Quick start

(see self-contained example)

We require a CAS file or cassis.Cas object that contains a text, e.g.

Die Fernuniversität in Hagen (Eigenschreibweise: FernUniversität) ist die erste und einzige staatliche Fernuniversität in Deutschland. Ihr Sitz befindet sich in Hagen in Nordrhein-Westfalen. 
Nach Angaben des Statistischen Bundesamtes war sie, ohne Berücksichtigung von Akademie- und Weiterbildungsstudierenden, mit über 76.000 Studierenden im Wintersemester 2016/2017[3] die größte deutsche Universität.[4]

The CAS is based on a type system file or cassis.TypeSystem object and specifies annotation types, e.g.

de.tudarmstadt.ukp.dkpro.core.api.ner.type.NamedEntity

To transform the text into an HTML string with underlined annotations of type NamedEntity, we run:

from cas_visualizer.visualizer import SpanVisualizer

cas = '../data/hagen.txt.xmi'
ts = '../data/TypeSystem.xml'

span_vis = SpanVisualizer(ts)

span_vis.add_type(name='NamedEntity')

html = span_vis.visualize(cas)

Finally, in a browser we can then render the HTML string:

Before visualizing the CAS, you can switch to the highlighted style by calling:

span_vis.selected_span_type = "HIGHLIGHT"

---

How to publish a new version:

1. Increase the version number in pyproject.toml
2. Run poetry build
3. [Optional] If no token is configured:
   • Create an API-Token by visiting: https://pypi.org/manage/account/#api-tokens
   • Replace TOKEN with the string of the API-Token and run poetry config pypi-token.pypi TOKEN
4. Run poetry publish