cts textgroup and work metadata files

File placement

Following the Perseus CTS publication guidelines for metadata files documented and developed at Capitains, each directory must have both a valid cts textgroup file and a valid cts work file.

The files are identically named as __cts__.xml (__ = two underbars), but each are nested in different, appropriate locations.

So each working repository must have a data folder and within this, files are organized by author/textgroup and works. For example, the file named tlg0007.tlg069.perseus-grc2.xml, with the author/textgroup tlg0007 and the work tlg069 would be found here:

 myworkingrepo/data/tlg0007/tlg065/tlg0007.tlg069.perseus-grc2.xml

Each of the folders that contain 1) the author/textgroup (tlg0007) and 2) the work (tlg069) must have its own cts metadata file.

So the author/textgroup cts metadata file is here:

 myworkingrepo/data/tlg0007/__cts__.xml

and the work cts metadata file is here:

 myworkingrepo/data/tlg0007/tlg069/__cts__.xml

Content of the metadata files

The structure of these files follow the guidelines linked above, but may be unclear how much or how little data should be included in the content of these files.

Author/Textgroup Metadata

A standard canonical name is preferred. This is typically informed by the Perseus Catalog records or derivative pick lists and/or other canonical sources. Metadata files should be as succinct as practical and only include additional information where needed to alleviate any confusion. As these files inform the software presentation, brevity and clarity are key.

Alternate names, transliterations, languages, various epithets, and dates are not recorded here, as it is expected that the Catalog will supplement this information, either via links or other tools. Any questions about the preferred name should be referred to librarians or editors for the appropriate collection(s).

In this example, Plutarch is the author:

 <ti:groupname xml:lang="eng">Plutarch</ti:groupname>

The full author/textgroup metadata in this example may be found here.

Work Metadata

Work metadata has additional elements not found in the author/textgroup metadata, but similar principles apply.

This file informs how the work is described in the reading environment. We do not incorporate multiple alternate titles in various languages or long descriptive titles. As noted, the Perseus Catalog does include this information.

Current practice is to include only one top level title in the work metadata file. (Note: prior to standardization of this work, some alternate titles were included.) As above, this title corresponds to the canonical title used in the Perseus Catalog or derivative pick lists and/or other canonical sources. If there is no existing canonical title, then one should be chosen in consultation with the librarian(s) and editors.

Take care to note the language of the title, if it is not in English.

In the example urn:cts:greekLit:tlg0007.tlg069, the canonical title as catalogued is:

 <ti:title xml:lang="lat">De recta ratione audiendi</ti:title>

As of this writing, the Perseus Greek collection contains one primary language edition and two translations of this work. For each of these, the metadata file will have a label and a description.

For the edition or translation label, current practice is to use the work title as it appears in print. So the label is functionally equivalent to the title, but reflects the print editor or translator's preference.

The description should succinctly summarize the bibliographic information for the print source of the edition or translation. It is similar to a print bibliography and should follow the source description found in the file header.

In this example, the Greek edition of the work has the following label and description:

   <ti:label xml:lang="grc">Περὶ τοῦ ἀκούειν.</ti:label>
   <ti:description xml:lang="mul">Plutarch. Plutarchi Chaeronensis Moralia, Vol I. Vernardakēs, Grēgorios N., editor.  
        Leipzig: Teubner, 1888.</ti:description>

The label is the Greek title as it appears in print, with the appropriate language attribute.

The description follows the general format:

Short_author_name. Print_edition_title, volume_#_(where applicable)_. Editor/Translator_Last, Editor/translator_First, role. Publication_city: Publisher, Publication_Year.

Include the appropriate language attribute here as well. In most cases, the description will include multiple languages ("mul") as it will generally have a mixture of English and Latin.

The short author name is meant to assist the reader; it is not intended to be the canonical author name. So, for example, "Cicero" is better than "M. Tullius Cicero" in this application. The short author name should be followed by a period.

The print edition title. Some book titles can be quite long. Include only the information necessary for the reader to quickly cross reference the edition in question. Multiple subtitles are not desirable. Not all editions will have a volume, but this should be noted in the case of a multivolume source. Arabic numerals may be used instead of Roman numerals, but whatever the print uses is acceptable. (Both "Vol. 8" and "Vol. VIII" may be used). The print edition title should be followed by a period. If a volume is noted, then the title will be followed by a comma and the volume information should end with a period.

The editor or translator's name, should be presented last name, first name and follow the information in WorldCat. In general, full names, when known, are used in place of initials to avoid confusion. The role will be either editor or translator, all lowercase letters. Follow the punctuation in the presentation: Last_name, First_name, role. Role will be followed by a period.

Publication information follows a print bibliography with City: Publisher, Year.

The full work metadata in this example may be found here.