Inscription corpora
Source catalogues, identifiers, maps and distribution over inscription records.
Data & computation数据与计算
Loading corpus descriptions… · 正在加载语料说明…
Read & compare阅读与比较
Search catalogues检索目录
Edit & analyse编辑与分析
Collections & tools合集与工具
Filter names and descriptions here. Search edition wording in Text search; source catalogues retain their own records.此处按名称与说明筛选。版本措辞请用“文本检索”；各来源目录保留自身记录。
Browse collections浏览合集
Papyri & ancient documents
Documentary and literary papyrus resources, with their source roles distinguished.
Editions & document dossiers
Source-qualified editions, reading views and associated case studies.
Inscribed objects
Inscribed objects, models, stamps, milestones and material support records.
Historical contexts
Related archaeological, geographical and historical work with mixed evidence.
Tools & methods
Encoding, querying, restoration candidates and computational methods.
Teaching & exercises
Courses, exercises, lectures and source-aware student work.
Reference & scholarship
Bibliography, historical copying, source versions and comparative readers.
Papyrological databases纸草学数据库
Edited texts, object catalogues, translations, images and bibliography retain separate roles. Local availability and text admission are recorded on each entry. · 区分版本正文、器物目录、译文、影像与书目；各项注明本地可用状态及文本准入范围。
Download a corpus database for offline full-text queries, or original XML for editorial analysis. Each archive preserves a frozen local snapshot and its own source licences; it does not claim current upstream coverage. Search results are edition records; shared identifiers do not collapse them into objects.下载语料库以离线检索全文，或下载原始 XML 进行版本分析。各文件保留本地快照及自身来源许可，不表示涵盖上游最新版本。检索结果按版本记录计；共有标识符不使之自动合并为器物。
| Corpus · 语料库 | Query data · 检索数据 | Sources · 来源 | Licence · 许可 |
|---|
Snapshot coverage & checksums快照范围与校验和
Offline full-text query离线全文检索
Download query.py and a corpus database, then decompress it. Python 3 with SQLite FTS5 is required. The query helper writes JSONL or CSV and keeps the edition and source hash on every record.
gzip -dk texts.sqlite.gz python3 query.py texts.sqlite '"dis manibus"' --layer reading --format jsonl python3 query.py texts.sqlite 'imperator -augustus' --format csv
The offline helper uses the same word normalization and quoted-phrase rules as browser text search. It has no browser date or language filters; exported record JSON retains those fields for subsequent filtering. CSV escapes formula-leading text for spreadsheet safety. JSON retains unmodified values.
This World centres on inscriptions, ancient documents and papyrus witnesses. Handwritten papyri can be manuscripts; literary papyri and later manuscript transmission therefore retain distinct roles. A modern edition or historical copy is not direct inspection of an original object.本世界以铭文、古代文书与纸草见证为核心。手写纸草本身亦可属于抄本，故区分文学纸草与后世抄本传承。现代版本或历史摹本不等于对原物的直接检视。
Sources & methods来源与方法
EpiDoc XML remains the edition source. Searchable text is a documented derivative. Translations, apparatus, metadata, teaching notes and restoration proposals have distinct roles and are excluded from this original-text index.EpiDoc XML 保持为版本来源。可检索文字是有说明的派生文本。译文、校勘记、元数据、教学笔记与修复候选各有不同作用，不收入此原文索引。
Edition reading selects the lemma and edited reading where encoded and includes supplements and abbreviation expansions. Exclude supplements & expansions removes encoded supplied/ex content and prefers orig/sic/abbr where present. It still contains uncertain or erased readings and is not a diplomatic transcription of surviving letters. Compare the XML and the physical witness before making such a claim.“编辑读法”采用编码中的主读法与编辑文本，含补阙及缩写展开。“排除补阙与缩写展开”去除 supplied/ex 内容，并在可用时优先采用 orig/sic/abbr。它仍含不确定或被涂抹的读法，不是对实存字母的外交式转写；作此判断前须比较 XML 与实物见证。
Origin, findspot and repository are separate fields. Dates retain their source interval and qualifiers; no point date is guessed. Catalogue identifiers relate records and editions; an archaeological object, text, publication and modern record are not treated as the same entity.制作地、出土地与馆藏分别记录。年代保留来源区间及限定，不猜测单一日期。目录标识符用于关联记录与版本；考古器物、文本、出版物与现代记录不视为同一实体。
The project directory connects existing Magalia builds. Navigation coverage does not establish shared corpus coverage, record identity, complete excavation context, or equivalent search capabilities.项目目录连接既有 Magalia 构建。导航覆盖不证明语料覆盖相同、记录同一、考古语境完整或检索能力相等。
Analytical lemmatized derivatives remain in source XML archives and are excluded from text queries. Representations marked transliteration, reconstruction or PHI are labelled and excluded from the conservative layer; primary representations follow their encoded markup. · 词形还原等分析性派生文本保留于来源 XML，不收入文本检索；标为 transliteration、reconstruction 或 PHI 的文本表示另行标识，且不参加保守层检索；primary 表示按来源已编码的编辑操作处理。
Only admitted, explicitly licensed source editions enter this snapshot. Sources held for publication review remain excluded; their original XML is preserved unchanged.此快照仅收入已准入且许可明确的来源版本。待公开审查的来源暂不收入；原始 XML 保持不变。
Reference systems & standards参考系统与标准
- EpiDoc Guidelines
- Papyri.info documentation
- I.Sicily open data and use
- FAIR Epigraphy
- IIIF Presentation API 3
Magalia EpiDoc guide · Bibliography · Matrix Hub reference material