The corpus in numbers
from catalog.csv at build timeDerived layers, counted on their own: L1 and L2 are generated by code from the witnesses, one file per witness file, so they are not counted as documents of the corpus (ruling CNT-1).
L0.5 (Egyptian transliteration with editorial symbols, the step before L1/L2): 2,438 documents, 1,101,453 words — a working layer, not counted in the corpus or in L1/L2.
Compositions are works (ruling WORK-1: Joshua, a homily, the Difnar), each read in one or more witnesses — they are not documents and are not added into any document count. The corpus holds 339 compositions; 318 are served here, and 21 have no share-safe witness. “Texts with English” counts the texts served with an English column actually lined up beside the text.
Catalogue rows by layer (25,140 rows, one per layer file): L0 witness 11,805 L1 Cairene 5,192 L2 Kemetic 5,192 L0.5 transliteration 2,438 L0 witness (OCR) 475 L2 Kemetic draft 38
Three layers, each generated by code from the one above it: L0 witness (the text as its edition has it), L1 Cairene (standardized Greco-Bohairic spelling and grammar), L2 Kemetic (the Kemetic alphabet). Corrections go into rules, never into output files.
Scholars
- Browse the library by genre and work, or walk the scroll wall
- Read any text with its layers side by side — e.g. Bohairic Gospel of Mark 1 (Digital Edition) — bohairic.mark
- Search the Coptic, Greek, Kemetic and English layers, one genre at a time
- The Images: 1,028 engraved plates of Egypt, public domain, by collection and volume
- Every page carries its source, licence, attribution and a citation — licences
Learners
- Read a text in the witness, in Cairene and in Kemetic, line by line, with English where it exists
- The Images: 1,028 engraved plates of Egypt, public domain, by collection and volume
- Guides for readers of Greco-Bohairic and for English speakers into Kemetic planned
- The children's guides, the rulings game and the Kemetic keyboard planned
Teachers
- Parallel texts to project: switch the layers you want on each text page
- Grammar tables and rulings with their corpus evidence planned
- What may be copied, printed or projected: licences by document
What is served openly
Six share classes (LICENCES.md §0/§11). Only public and attribution are served here; the TLA
sentence corpora appear only as our derived layers (ruling F7-3). The reading room for the other classes waits for the permission replies.
| share class | catalogue rows | on this site | why |
|---|---|---|---|
public | 7,479 | served openly | public domain or an open grant; the edition is still named |
attribution | 14,748 | served openly | with the attribution line kept on every page and every hit |
restricted | 1,926 | not on this site | terms allow it only non-commercially or per reader — for the reading room, which waits (G1-4) |
internal-unstated | 9 | not on this site | nobody has stated terms; the permission requests are outstanding |
internal-forbidden | 327 | never leaves the machine | terms exist and forbid redistribution |
unverified | 651 | never leaves the machine | no licence found and no edition named — not until ruled |