]> Thutmose II MARCXML to TEI Header translation Española, New Mexico Black Mesa Technologies LLC 2009

No source; created in electronic form.

 Thutmose II MARC XML to TEI header translation C. M. Sperberg-McQueen, Black Mesa Technologies LLC 21 December 2010 rev. &date.last.touched;

Thutmose II translates from MARCXML records to TEI headers. The II indicates that this is the second version of the program; an earlier version of the program (Thutmose I) is also available. When complete (which should be any day now), Thutmose II will handle all of the fields described in the TEI Library SIG's best practices document.

Usage

This document assumes you have an XSLT 1.0 processor and know how to invoke it on your MARCXML data, and how to pass parameters to the stylesheet processor.

Run-time parameters

The most important parameters are these: Specifies which result element to use for each MARC record in the input: TEI means to produce a complete tadpole TEI document, containing a header with a rudimentary text structure. teiHeader means to produce just the header, without the TEI wrapper element and rudimentary text. If the input contains more than one MARC record, the set of resulting elements (whether TEI elements or teiHeader elements) will be wrapped in a teiCorpus element regardless of the value of result. Supplies the URI (or local file name) of an XML document with configuration information specifying how information from the MARC record is to be mapped into the TEI header user-supplied text to be inserted at appropriate places in the output. For example, the MARC record is assuemd to describe the source being digitized, not the TEI document itself; the publication statement for the TEI document will typically identify the project which is creating the TEI document; appropriate XML should be provided in the configuration file. Defaults to thutmose.config.xml.

Additional parameters

Additional parameters include: Use the value specified both as the value of the id attribute on the TEI element (if one is generated) and as the value of an idno element in the publication statement. The year in which the TEI document is being published. If not supplied as a parameter, this is taken from the configuration file. If not present there, Thutmose II tries to guess; sometimes it may guess right. Specifies (if the value is 1 or true) that data from the MARC record should be wrapped in TEI seg elements with type attributes indicating which MARC field and subfield the data came from. Useful for debugging; probably clutters the output too much for normal use.

Getting Thutmose II

Just download what you need: mt2.xsl the stylesheet itself thutmose.config.xml the default configuration file userdoc.xml this document (in TEI Lite) progdoc.xml a document aimed at people who want to modify or maintain the stylesheet (in TEI Lite)

Many XSLT processors will be able to retrieve the stylesheets from the Web, so you may not need to download them at all. But if you're going to using Thutmose a lot, it's probably better network citizenship to download a copy and check here periodically for updates.

Plans for the future

This is the second version of Thutmose.

Follow-on projects are expected to: improve the data beautification options provide some record of changes made in the course of beautification, so that it can be reviewed and the data can be edited if necessary; changes will be recorded either inline (in processing instructions) or in a log emitted as a side effect of transformation map from TEI headers to MARC (currently this has lower priority)

Known gaps, bugs, and shortcomings

In the default configuration, Thutmose II handles, by design, only the mappings specified in the Best Practices document, and only MARC 21 fields not withdrawn, marked OBSOLETE, or otherwise deprecated.

If you have MARC records with fields no longer present in MARC 21 (say, they have 440 fields for series titles), you can add mappings for them by customizing the configuration file.

Features planned for Thutmose II

All of the following features are expected to be part of Thutmose II. (That is, I don't plan to declare Thutmose II complete until these are all done. The reader may think of this as a to-do list for the programmer.)

The respStmt elements being generated by MARC fields 100 and 700 are, to put it gently, sub-optimal. That is to say, they are very very bad.

(23 December 2010: this should be fixed in the medium-term future.)

MARC fields 100 and 700 should tag their contents (or: the relevant parts of their contents) as personal names.

(30 December 2010: this should be fixed in the medium-term future.)

MARC fields 110 and 720 generate author, editor, or respStmt elements as appropriate. Their content should also be tagged as a corporate name.

(21 December 2010: this should be fixed in the medium-term future.)

MARC fields 050-099 should generate idno, classDecl, taxonomy, and classCode elements as appropriate. This requires some close reading of the MARC 21 documentation which has not yet been done; currently these MARC fields are ignored.

(21 December 2010: this should be fixed in the medium-term future.)

Other

The following items all describe features and functionality that will not be part of the first complete version of Thutmose II; they may be added as and when I have time and inspiration.

The user interface to the online demo version of Thutmose should be refined to make it easier to see what is going on. One way to do this would be to provide both a read-only view (with formatting and syntax coloring) and an editable view using text widgets (as in the current interface), with buttons to switch between views.

MARC fields 100 and 700 generate author, editor, or respStmt elements as appropriate. Some subfields, however, don't belong in those elements and should be filtered out. User options should be available (through the configuration file) to control this filtering.

The beautification (punctuation-stripping) routines which are part of Thutmose I are not yet integrated into Thutmose II.