Who I am
I am a software developer with a deep interest in Coptic language and heritage. Like many learners outside Egypt, I came to the language through church liturgy, reading, and curiosity about the last stage of Ancient Egyptian — not through a formal classroom alone.
I am not a professional linguist or a clergy member. I build tools because I needed them myself: faster lookup, readable Unicode, and pronunciation help while studying texts that still use legacy fonts and scattered reference books.
Why this is personal
Coptic belongs to a living community — in Egypt and across a global diaspora. I wanted work that serves that community openly rather than locking dictionaries and converters behind closed formats or one-off scripts only I could run.
Why I started
It began with frustration at ordinary tasks. Looking up a word meant flipping through PDF scans or guessing how a headword was spelled in dictionary Latin. Converting Antonios-font text to Unicode was manual and error-prone. Pronunciation rules were spread across notes, books, and audio with no single place to practice.
The classical dictionaries — especially Crum — are extraordinary, but they were not built for how we search today. I started extracting and structuring entries so I could search by Coptic word, transliteration, or meaning in one place.
The work behind it
None of this appeared at once. It was built in stages over months, where each step was only possible because of the one before it. What looks like a few tools on a page is really several distinct pieces of work — each needing both software and a real understanding of the language.
Mapping the old font to Unicode
Most Coptic text online was never real Unicode — it was typed in legacy Antonios fonts, where ordinary Latin keys are simply drawn as Coptic letters. That text breaks the moment you don't have the exact font. To fix it I first had to know the full mapping, so I built an app that displays every glyph of the Antonios font and, using an on-screen keyboard, matched each one by hand to its true Coptic Unicode character — including the hard cases: combining marks, the overline, the jenkim, and punctuation. That verified mapping became the coptic-antonios-unicode converter.
Teaching a machine to pronounce Coptic
Bohairic pronunciation is not a simple letter-for-sound swap — some letters change depending on their neighbours, and certain combinations make a single sound. I encoded these rules into an English transliterator, testing against real words and refining the exceptions one by one.
Then again, in Arabic
Because the largest Coptic community reads Arabic, I added a second target: Coptic Unicode → Arabic phonetics. Arabic has its own sound inventory and conventions, so the rules were rewritten from scratch — but both now live behind one simple, validated tool: coptic-pronounce.
Turning a printed dictionary into data
The classical dictionaries exist as PDF scans laid out for print, not for searching. I used the Adobe PDF Extract API to convert the pages into structured data, then wrote special-purpose parsers to rebuild real dictionary entries — headwords, senses, glosses, variants, cross-references — that the print layout had scattered across columns and lines. The output is ~205 searchable JSON pages with indexes by word, transliteration, and meaning.
Adding what the book never had
The original PDF contained no Unicode and no transliteration. Both had to be generated — and could only be, because the earlier tools already existed. I ran the Antonios→Unicode mapping and the pronunciation engines across every entry, so the dataset now carries information the original dictionary never had. It is not a scan of a dictionary; it is a new, enriched dataset derived from one.
Why the order matters
Each step made the next one possible. The enriched dictionary could only be built because the Unicode mapping and the pronunciation engines already existed and were correct. The whole thing needs two kinds of knowledge at once — software engineering and the Coptic language itself.
This site ties those pieces together under /app: dictionary search, Unicode conversion, and pronunciation practice in interfaces meant for daily use. The two libraries — coptic-antonios-unicode and coptic-pronounce — are open-sourced so others can reuse them in their own apps and research.
What I hope for
I hope students, researchers, and families in the diaspora spend less time fighting tools and more time with the language itself — hymns, manuscripts, conversation, and prayer. I hope corrections and contributions from readers make the data more accurate over time.
If you spot an error, have a manuscript source to suggest, or want to help improve the codebase, please get in touch.
For the language itself and our mission, see About Coptic.