A corpus of citable texts must include:
hocuspocus represents a corpus with the Corpus class.
A typical situation is that editors work with files in a local file system. Editions of texts and the cataloging information are both typically structured in XML. Editions may use any XML vocabulary in any namespace (or in no namespace). XML files of cataloging information validate against the TextInventory.rng schema defined in the cite library.
hocuspocus maintains the configuration data needed to work with local files in a CitationConfigurationFileReader. This includes information about where to find files in the file system, and how the XML markup of individual files is related to canonical citation schemes. The constructor for a Corpus object therefore includes a local XML file with citation configuration data, as well as an XML TextInventory and a base directory for finding XML editions.
The Corpus object includes two specially useful high-level methods. The tabulateRepository loops through all texts in a corpus and creates tabular files from local XML files. The turtleizeRepository method first tabulates all texts in a corpus, then composes an OHCO-2 equivalent represenation in TTL.
To convert a corpus of local XML files to tabular files, the Corpus object uses a Tabulator object. Its tabulateFile method can convert a single XML file to tabular format.
To convert tabular files to RDF, the Corpus object uses a CtsTtl object. Its turtleizeFile method converts a single file of tabular data to RDF.