umschriftpy

A Python 3 library for dealing with Egyptological transliteration/transcription.

---

NB If you are looking for an end-user tool to convert Egyptological transliteration/transcription to Unicode, please visit the Transliteration to Unicode Converter.

---

Description

This library is made for standardising Egyptological transliteration/transcription with the aim of processing data from digital projects with different approaches to Egyptological transliteration/transcription. umschriftpy should be able to import Egyptological transliteration/transcription entered in different Unicode flavours, as well as in the conventional fonts Umschrift_TTn, Transliteration and Trlit_CG Times, and using different Egyptological conventions. One particular application for which it was developed is the matching of entries in different databases. The Leiden Unified Transliteration/Transcription https://www.iae-egyptology.org/the-leiden-unified-transliteration serves as the basis for the normalisation.

It is intended to be ported to other common programming languages used in the development of Egyptological applications.

Usage

The package provides a class for working with Egyptological transliteration/transcription called UmschString and methods for creating objects of this class from strings. To create an UmschString object you can use one of the following methods:

• from_unicode loads Unicode-formatted strings
• from_umschrift_ttn loads strings formatted with Umschrift_TTn v3.0 font (https://wwwuser.gwdg.de/~lingaeg/lingaeg-stylesheet.htm)
• from_trlit_cg_times_2023 loads strings formatted with the 2023 version of the Trlit_CG Times font (https://dmd.wepwawet.nl/fonts.htm https://oeb.griffith.ox.ac.uk/fonts.aspx)
• from_trlit_cg_times loads strings formatted with the pre-2023 version of the Trlit_CG Times font
• from_transliteration loads strings formatted with Transliteration font (CCER), downloadable from the link above All of these methods can be modified using the optional UmImport.S_FOR_Z flag, which imports s and S as z and Z.

The UmschString objects have export methods for outputting strings and support some of the Python string methods (upper, lower, index, find, replace, endswith, startswith) Data can be cleaned up using the filter method, with filtering options set by flags:

UmFilter.MORPH removes morphological markers . : · UmFilter.SUFF_PRON removes suffix pronoun separators ⸗ UmFilter.BRACKETS removes brackets ⸢ ⸣ ⟨ ⟩ ( ) [ ] < > { } | UmFilter.PUNCT removes punctuation ? ! " , . UmFilter.CLEAN removes all of the above UmFilter.REPLACE_Z uses s for z and S for Z UmFilter.REPLACE_I_WITH_DIAERESIS uses y for ï UmFilter.REPLACE_INVERTED_BREVES uses ꞽ for i̯, w for u̯ UmFilter.REPLACE_UNCERTAIN_CONSONANT uses ꜣ for ꜥ UmFilter.REPLACE_ALL replaces all the three of the above UmFilter.DIGITS removes all digits UmFilter.FACULTATIVE removes parentheses and all signs enclosed in parentheses UmFilter.LOWER converts all transliteration/transcription to lower case UmFilter.HYPHENS replaces all hyphens with spaces

Data can be normally to string using the to_unicode method. By default this method uses the Leiden Unified Transliteration/Transcription, but deviations can be set up by optional flags: UmExport.K_WITH_DOT uses ḳ and Ḳ for q and Q UmExport.J_FOR_YOD uses j and J for ꞽ and Ꞽ UmExport.JJ_FOR_DOUBLE_YOD uses jj and Jj for y and Y as well as for ï and Ï UmExport.REPLACE_I_WITH_DIAERESIS uses y and Y for ï and Ï UmExport.Z_FOR_S_AND_S_FOR_S_ACUTE uses s / S for z / Z and ś / Ś for s / S (Wörterbuch-style transliteration/transcription) Flags can be combined using the bitwise OR | operator

Besides, data can be output using the to_pseudo method. It produces pseudo-coded Unicode strings, which can be stored, for example, in a database TEXT field and used for (case-sensitive) binary-based sorting and string comparison of the original transliteration/transcription strings (tested with SQLite and MariaDB/MySQL). Pseudo-coded Unicode strings can be loaded into UmschString objects using the additional from_pseudo method. For an example, see the test_pseudo() method in test_pytest.py. Alternatively, the to_transliteration exports data to ASCII-codes, which are correctly displayed using the Transliteration font (CCER).

Examples:

from umschriftpy import *

#       CONVERSION

# imports a Trlit_CG Times string
uma = from_trlit_cg_times(
    "Htp dj nsw Wp-WAwt nb tA Dsr dj=f prt-xrw (m) t Hnqt")
uma_lower = uma.lower()  # converts it to lower cases
uma_replaced = uma_lower.replace(from_trlit_cg_times(
    "nsw"), from_trlit_cg_times("njswt"))  # replaces a word
print(uma_replaced.to_unicode())  # exports to Unicode
# >>> ḥtp dꞽ nꞽswt wp-wꜣwt nb tꜣ ḏsr dꞽ⸗f prt-ḫrw (m) t ḥnqt
# exports to Unicode with a special option (ḳ instead of q)
print(uma_replaced.to_unicode(UmExport.K_WITH_DOT))
# >>> ḥtp dꞽ nꞽswt wp-wꜣwt nb tꜣ ḏsr dꞽ⸗f prt-ḫrw (m) t ḥnḳt

#       CLEANING UP STRINGS FOR COMPARISON

a = from_umschrift_ttn("O#.tj-o-jj#")  # imports an Umschrift_TTn string
print(a.to_unicode())
# >>> Ḥꜣ.tꞽ-ꜥ-yꜣ
b = from_unicode("ḥꜣtꞽ ꜥ yꜣ")  # imports an Unicode string
print(b.to_unicode())
# >>> ḥꜣtꞽ ꜥ yꜣ
# the strings are different because of different punctuation, lowercase and uppercase letters
print(a == b)
# >>> False
a_filtered = a.filter(UmFilter.CLEAN | UmFilter.HYPHENS | UmFilter.LOWER)
b_filtered = b.filter(UmFilter.CLEAN | UmFilter.HYPHENS |
                      UmFilter.LOWER)  # filters strings
print(a_filtered == b_filtered)  # filtered strings are now equal
# >>> True
print(a_filtered.to_unicode())
# >>> ḥꜣtꞽ ꜥ yꜣ

#       SORTING

# inputting several names as a list
name_list = [from_transliteration("nfr-Htp"), from_transliteration(
    "aA-ptH"), from_transliteration("DHw.tj-nfr"), from_transliteration("Aw-jb")]
# filtering list values
name_list_filtered = [item.filter(UmFilter.CLEAN | UmFilter.LOWER)
                      for item in name_list]
# using a standard Python function to sort items
name_list_sorted = sorted(name_list_filtered)
# outputting the sorted list as Unicode
print([item.to_unicode() for item in name_list_sorted])
# >>> ['ꜣw-ꞽb', 'ꜥꜣ-ptḥ', 'nfr-ḥtp', 'ḏḥwtꞽ-nfr']

Why a transliteration/transcription library?

When dealing with Egyptological transliteration/transcription in database projects, one is faced with several difficulties. While some projects use Unicode for transliteration/transcription, others still use legacy fonts. There are also different variants of Unicode-based encoding (different codepoints for aleph, ayin and yod), not to mention different systems of transliteration traditionally used in the discipline (ꞽ or j; q or ḳ; ṯ/d/ḏ0 or č/ṭ/č̣), which makes it difficult to match transliterations across databases. Database systems and standard programming language libraries do not natively allow sorting of Egyptological transliteration/transcription values. Irregularities sometimes occur when programmably comparing Egyptological transliterations/transcriptions. I think that one possible way to streamline the processing of Egyptological transliteration/transcription in digital projects is to develop dedicated libraries for this particular task. Ideally, these libraries should be ported to a variety of programming languages used in digital Egyptology to serve as building elements for different applications.

Disclaimer and acknowledgements

The current preliminary version is merely a proof of concept.

The package is being developed as part of the project “Altägyptische Titel in amtlichen und familiären Kontexten, 2055-1352 v. Chr.”, funded by the Fritz Thyssen Foundation.