-
Notifications
You must be signed in to change notification settings - Fork 0
Home
The first edition called here edition A is available for download from Małopolska Digital Library in the unbundled DjVu format. After removing the covers and empty pages and converting it to the bundled form it is available in this repository as Zaborowski-MBC.djvu.
The second edition called here edition B is available in Polona as graphic files served according to IIIF standard which can be downloaded individually as JPG or TIFF and as PDF as the whole. They have been downloaded as JPG and converted to DjVu with didjvu (title page and empty pages were also skipped). The result is available in the repository as Zaborowski-Polona.djvu.
Both documents have been supplemented with the metadata and outlines.
The best way to view them is to download them and use djview.
The standard of character encoding is Unicode; the selected so called Private Use Area characters are used according to the current specification of Medieval Unicode Font Initiative.
Interpreting printed type as Unicode characters is not always obvious. Some problems of this type has been discussed in Polish in the preprint available at https://www.researchgate.net/publication/341930612.
Histograms of the characters used in abbreviation in each edition are provided in the repository. The have been created with the unihistext program.
The tentative transciption of the texts has been made with Transkribus. The results has been used as the basis for indexes, which were later verified and extensively modified. These changes has not been applied yet to the texts stored in the Transkribus system.
Th program for Debian GNU/Linux and Windows is available at https://bitbucket.org/mrudolf/djview-poliqarp/.
Unfortunately the online help doesn't cover the most recent features of the program. Their description is available in Polish in a presentation and a paper.
Indexes are files in a variant of the CVS format using semicolons as delimiters, encoded in UTF-8; they have the extension 'cvs'. Every line consists of three or four fields:
- entry used for sorting and incremental search,
- reference to the relevant image fragment in the form accepted by djview viewer,
- description: a text displayed for the current entry in a small window,
- optional comment displayed after the entry.
The program allows to concatenate indexes easily. The entries can be displayed in several orders:
- file order,
- alphabetic word by word,
- alphabetic letter by letter,
- a tergo (with some limitations).
Pressing the middle mouse button in the left side graphic panel of djview4poliqarp should open djview displaying the relevant page; this is useful as djview has some features absent in djview4poliqarp, such as displaying outline and metadata. However for this to work the references in the index must contain absolute paths to the files while the repository provides the indexes with the relative paths only; fortunately the user can change them easily to absolute ones by a simple global replacement in a text editor.
The primary indexes are the indexes of abbreviations listed in the order of their occurences in the treatise, supplemented by some auxiliary entries. Hence the basic entry line has the following structure:
- abbreviation, e.g. 'mãibꝰ' (you need an appropriate font to display the example correctly),
- a local reference, e.g. 'Zaborowski_MBC.djvu?djvuopts=&highlight=561,954,133,58&page=1'; reference of this form use default color for highlighting, this can be changed by adding additional djview parameter (for the references to work djview4poliqarp has to be started from the directory containing the index),
- description, usually empty,
- the abbreviated word, proceeded by REFERENCE MARK for more distinctive display, e.g. '※ manibus'.
Additional entries are of two kinds.
First of all there are entries describing words which are not abbreviations but are interesting for other reasons; in particular, they document the usage of LATIN SMALL LETTER ET as a final 'm'.
Secondly, they are entries allowing, when displaying an index in the file order, to move quickly to a specific page or, when both A and B indexes are diplayed together, to the beginning of an edition.
The indexes are named respectively ZaborowskiA.cvs and ZaborowskiB.csv.
In the secondary indexes the fields 1 and 4 are exchanged. They are generated from the primary ones with a 'sed' one-liner. They are named respectively ZaborowskiAi.cvs and ZaborowskiBi.csv ('i" meaning 'inverted').
Loading both indexes and sorting the joined index in an alphabetic order allows to compare how the words were abbreviated in the A and B editions.