Hkoll

Hkoll was developed to collate several text versions with one ›main‹ version and calculate an apparatus criticus from the differences. Deviations between the textus constitutus (called »model«) and one or more transcribed witnesses (called »images«) are classified as addition, omission, transposition or – if none of the former applies – plain variance.

Running Hkoll implies the following steps:

1. Parsing model and images (these are provided to the software as separate XML/YAML files and must follow some specific syntax rules, cf. Data Format).
2. Normalising the image readings in order to avoid trivial variants from different spelling.
3. Aligning and collating the model and images: This is the core of Hkoll’s algorithm, but in a future release it is planned to offer also the use of CollateX (https://collatex.net/) for the alignment.
4. Integrating standoff information, i.e. any kind of information linked to a specific part of the model (e.g. philological or other commentary, translation, apparatus fontium).
5. Formating the result as one comprehensive (TEI) XML file which may serve as input to further processing, e.g. to create HTML or PDF versions.

The main difference to other collation tools is that Hkoll supports a highly automated workflow. Indeed, the idea is that Hkoll calculates the complete edition (taken as information, not as a specific representation, cf. Thomas Stäcker: A digital edition is not visible) from the given input – model, images, normalisation rules, standoff – without any human interception afterwards. All decisions taken by the editors are instead included into the input files; Hkoll then derives the consequences.

The advantage of such an automated workflow is that it enables additions, corrections, and revisions at any time without the need of manually adjusting the apparatus criticus. This is done by Hkoll in short time (seconds to minutes).

Demo: https://github.com/omsinf/hkoll-demo

Paper: Oliver Schrader, Sebastian Gassmann, Tessa Gengnagel, "Automatisierung und CI/CD in digitalen Editionen am Beispiel des Conciliator," in: Konferenzabstracts der DHd2024 Quo Vadis DH?, Passau, 21 February 2024, https://doi.org/10.5281/zenodo.10698406.