Text-Fabric v8.3.4: Text-Fabric with a new display algorithm

Dirk Roorda (Developer)

Research output: Non-textual formSoftwareScientific


In Text-Fabric we want to display pieces of corpus material in insightful ways.

We have implemented two types of displaying:

plain: almost as the plain text of nodes, but with optional in-line formatting
pretty: almost as a tree-bank view, but for the fact that the text objects are not merely trees, but graphs.
In both types of display it is possible to optionally show node numbers to the relevant pieces of text, and to highlight nodes.

In addition to highlighting, the displays can be tweaked by passing a fair number of options, in order to show of hide features of nodes, call in additional graphics, show or hide intermediate levels, etc.

Whatever we want to display, we have to display in HTML, which is basically a hierarchically organized set of presentation elements.

But a node and its constellation of relevant neighbourhood nodes does not have a hierarchical structure, in general, that is.

The unravel algorithm solves the problem of turning a node and its associated piece of the textual graph into a tree of node fragments in such a way that the order of the slots is preserved.
Original languageEnglish
Media of outputsource code/data file (online)
Publication statusPublished - 26 Jun 2020


  • annotation
  • corpus analysis
  • text-fabric
  • search engine
  • python
  • text modeling
  • hebrew
  • cuneiform
  • arabic
  • greek


Dive into the research topics of 'Text-Fabric v8.3.4: Text-Fabric with a new display algorithm'. Together they form a unique fingerprint.

Cite this