Revise and properly document corpus.xml

[cmil]

At the DraCor Summit an extension of the corpus.xml has been discussed (see sample below). The function and structure of this document need to be properly documented in the guidelines.

Also see #44, #68, #87.

[lucagiovannini7]

beispiel_corpus_xml.xml

[cmil]

I added a slightly revised version of the sample file to #167.

[lb42]

The current corpus.xml file is not valid TEI, for two reasons, both easily fixed. If it is meant to be a valid TEI header, it must contain a <sourceDesc> element'; secondly, it is meant to be a valid <teiCorpus>, it must contain some element (typically a <text>) additional to the <teiHeader>.

I suggest that the <sourceDesc> should contain simply one or more <p> elements describing briefly how the corpus was derived, e.g. "Retagged from the Gutnberg HTML text and proof read against the page images of the original printed text as shown in Google books" or "Texts converted using ezdracor".

I also suggest that there is no need for the enclosing <teiCorpus> element: the root element for this document should simply be <teiHeader>, and the dracor schema could be modified to permit this element as root. An additional framework document containing a TEI corpus element could then embed/include this header along with one or more other TEI elements to form a complete and valid TEI document.

Alternatively, the corpus.xml file should not claim to be a TEI document, if its usage is specific to dracor-processing, and there is no intention of making it valid.