Data
The corpus as files, for work that does not fit in a browser.
- Texts
- 750
- Pages
- 96,093
- With an author
- 293
- With a year
- 363
The OCR is machine-generated and uncorrected, and the cataloguing is uneven — an author is recorded for about two texts in five and a year for about half. Absent fields mean unrecorded, not empty, so a filter never describes the texts it leaves out.
Per query
- Search results as CSV:
/search.csv?query=… - Concordance lines as TSV:
/concordance.tsv?q=… - Anything else, page by page, over the AI Suite
How to cite
Cite a work, or one page of it, at these addresses:
https://www.viet-dien.com/t/<item_key> https://www.viet-dien.com/t/<item_key>/<page>
These addresses are built from the library’s own identifier, not from a database row number, so a citation stays valid.
Reuse
The scans and records belong to the holding libraries, and only a small number of items state their reuse terms here. Where nothing is stated, treat it as unknown and ask the library. The OCR text and the catalogue compilation are offered for research use with attribution.