Southern Italian Angevine Chancery Papers in XML, the script: MaRa v2.0

Alfredo Cosco

alfredo[dot]cosco[at]gmail[dot]com

https://github.com/orazionelson

DOI

MaRa stands for: Marcatore dei Registri Angioini

GitHub repository: https://github.com/angioXML/MaRa2

This software is part of a project of University of Napoli "Federico II", that aims to build an XML archive, a database and a search tool for the Southern Italian Angevine Chancery Papers.

A short historical intro

During the World War II, at the end of September 1943, a German army troop retreating from southern Italy burned the Angevine chancery Archive of Naples Kingdom. The papers were hosted, for safety reasons, in a village, San Paolo Belsito, about 30km far from their building in Naples (then almost daily bombed by the Allies). The German retaliation destroyed thousands of folders containing parchments from the 13th to the 15th century.

After the war, the archivists of Archivio di Stato di Napoli, led by Riccardo Filangieri, went in search of the transcripts of the lost acts, administrative and political, which had been made over the centuries by archivists, scholars and historians. Then, they started to rebuild the archive in printed volumes. The reconstruction work continues today, it reached the number of 50 volumes and it covers a range of years from 1265 to 1434, under the main title: I Registri della Cancelleria Angioina.

Software Dev for Angevine Chancery Papers in XML

Since 2004, Prof. Roberto Delle Donne, from University of Napoli "Federico II", managed for a digital edition of the Angevine Chancery sources, with the aim to realize a research tool accessible through a web interface.

I was involved, as a student of Arts and Literature and worker in IT, at the early stages of the project.

Focusing on available options XML appeared as the most suitable format to store the data, then a DTD was defined by Dott. Manuela Schiano. According to the DTD each XML had to be structured in three parts:

The content was structured in: PROTOCOLLO, TESTO and ESCATOCOLLO, that are the three main portions of a medieval document. Inside the document content each part contains children defined with the names of diplomatic: INSCRIPTIO, DATATIO, SANCTIO...

As early transcripts where made by random documents, a series of difficulties arose:

Moreover, the frustration of editing by hands full sets documents repeating tags hundreds of times, or the time spent to debug them when transcribed by other editors, brought to a different approach:

let's start from raw .txt files and automate mark-up as much as possible.

Then, this target became the object of my degree thesis in Medieval History: Alfredo Cosco, Critica documentaria e nuove tecnologie. Problemi, tecniche e un caso di studio, Università degli Studi di Napoli "Federico II", aa 2007-2008.

MaRa v1 was a set of unordered PHP scripts with a minimal web interface. it was born to transform text transcripts (made in a standard format easier than XML) in well-formed XML.

MaRa v2 is its evolution and was used to realize a first heavy draft of our XML repository.

All the scripts were rewritten or nested in CodeIgniter, a good option to add features with little effort, extended by an HMVC CMF called No-CMS, really useful and comfortable to build a user interface around the already written functions.

All the functions now are in models/ and controllers/ of mara2/ module.

The new interface integrates a highly customizable text editor like CodeMirror in order to manage everything inside the browser.

There is a dedicated set of regex written for CodeMirror to highlight the text files with different colors: useful to check transcript formal errors; it's possible to see them in the file:

registridev/modules/mara2/assets/scripts/angevin.js

The DTD has been replaced by an XSchema.

And, that's the most important thing, the job is designed in a flow:

from the transcript import to the export in a set of xml valid documents eventually packed in a zip file.

MaRa2 workflow

The starting point is the panel in:

registridev/mara2/mara2 controller, here you have a TOC based on printed editions designed to manage the flow.

TOC Panel

You can create a registry and import the text file.

Then you can Manage the registry in 5 steps.

Steps manager ​

Few notes about the transcript standard

Each registry transcript is made by one .txt file, it contains different partitions, each partition contains from 1 to n documents.

Each partition is separated by the tag:

The first line after the tag is the partition title.

Then the documents.

Each document must have this structure:

The pagination follows the printed text, like this example:

So, an end transcript should look like this:

The STEPS

Step1a

This STEP splits the entire registry into partitions using the tag #PARTITION#.

Step1a Partitions

The highlighted text in the integrated editor helps to find possible transcript errors.

Step1a

Step1b

Same as before but it adds page numbers to each document

Step1b

Step2a

This is the core of the application, it inherits most of MaRa1 functions. The software splits the partitions in single documents and import them in 4 db tables:

(where XX is the registry id number)

At the top of the page there is an Info Panel that works as a first error control:

if the 3 numbers in the panel are the same everything is going well otherwise there's something wrong in the transcript

Each document is analyzed in the tabs at the middle the page.

Step2a

The software normalizes the most common errors appeared in editing using regular expressions. Sometimes it corrects them, in other cases it shows an error message.

Sample error: in the next image the document title has not been separated by the text.

Step2a Error

It is possible to correct the error in the integrated editor.

Step2a Error

Sample error: in the next image the software alerts for a leak of the page number.

Step2a Error

Sample correction: in the next image the software advises for an autocorrection.

Step2a Error Correction

Sample Corrections: in the next image the software advises you and correct the notation ibidem taking the info from the previous record.

Step2a Correction

The tab for the secondary sources, called Fonti, will split and normalize them.

See in the next image that the notation Ms. Bibl. Naz changes to Biblioteca Nazionale di Napoli, ms.

Step2a Secondary Sources It is possible to edit the transformation DB from the sub-tab Casistica di trasformazione

Step2a Transformations

Sometimes the split vision of the sources shows you that something is wrong, see the next image:

Step2a hand Correction

Once there are no more errors it is possible to click on the button Approva to release the partition.

This will lock the partition and will unlock the next one.

When all partitions are released, on the page appears the Crea XML that leads you to the next step. Step2a go to Step3a

Step3a

This step transforms the records into XML files.

Step3a Make XML

Step3b

The last step is dedicated to validating the XML. It is possible to validate the entire transcription or the single document.

Step3b Validate

Then at the bottom of this interface is possible to pack each generated registry in a zip file or create a TOC of them.

Step3b Zip

The workflow is strictly connected with the filesystem organization inside the directory:

the sub-directories:

At the end of the flow, XML files will be stored in 03_xml/.

Targets

Almost the half of the entire documents set edited in I Registri della Cancelleria Angioina has been transformed to XML using MaRa v1 and v2.

Moreover, the design of the project, bring more consciousness about how to:

design an editing infrastructure based on XML but dedicated to people that have to practice XML as less as possible.

The know-how and the files were transferred in the new project based on xquery and eXist-db.

Notes about the software installation for testing

MaRa2 is designed for a localhost use, moreover, its specifical features make its use for a different project a nonsense, anyway, its installation and use comes as common CMS in a LAMP environment:

The package in this repository contains a sample registry (R11) ready for the STEP2A, you can try to find and correct some errors then transform it in XML files.

References

I registri della Cancelleria Angioina, Napoli, since 1950

Riccardo Filangieri, L'Archivio di Stato di Napoli Durante la Seconda Guerra Mondiale, Napoli, 1946

Jenkinson, Hilary, Sir & Bell, H. E. (Henry Esmond) & British Committee on the Preservation and Restitution of Works of Art, Archives, and other Material in Enemy Hands (1947). Italian archives during the war and at its close. H.M.S.O, London

Commission Internationale de Diplomatique. Comité International des Sciences Historiques, Vocabulaire International de la Diplomatique, ed. M. M. CÀRCEL ORTÌ, València 1994 e1997 (2a ediz. corr.).

LAMP: https://en.wikipedia.org/wiki/LAMP

PHP: http://www.php.com

CodeIgniter: http://www.codeigniter.com

jQuery: https://jquery.com/

No-CMS: https://github.com/goFrendiAsgard/No-CMS

CodeMirror: https://codemirror.net/

xquery: https://www.w3.org/XML/Query/

eXist-db: http://exist-db.org