Scholaris glossary

The words

Anchor (ancla)

The exact place of a passage inside its document: the physical page and the printed folio, the second of a recording, the slide, the sheet and rows of a table, or the section and paragraph of a web page with its access date. Every citation is written from an anchor.

Autocite (autocita)

The function that reads your draft, finds a passage in your library for each claim and proposes the citation with its page; you accept or discard.

Notebook (cuaderno)

Markdown notes with cards of verified passages that can be checked again against the library.

Vector space (espacio vectorial)

The model, version and dimensions used to compute some vectors, e.g. "gemini-embedding-2@1536". An .spdf can carry several.

Record (ficha)

A document's metadata (authors, title, year, edition, publisher, DOI, ISBN…), each field with its provenance.

Folio

In early books, each numbered leaf, with its recto (r) and verso (v): "fol. 23v". In Scholaris, by extension, the page number printed on paper, as opposed to the page's position in the file.

Foliation (foliación)

How a book is numbered: by pages or leaves, in roman and arabic numerals, with jumps and unnumbered plates. Scholaris rebuilds it page by page.

Fragment (fragmento)

The passage that is searched and cited: a stretch of text with its start and end anchors.

Modernised spelling (grafía modernizada)

A copy of the text with spelling brought up to date ("dixo" → "dijo"), stored separately and used only for searching. The cited text always keeps its original spelling.

Hash (huella)

The SHA-256 digest of a file; two files with the same hash are the same file.

Imprenta (the press)

Scholaris's converter: it recognises each file's type and prepares it for reading (pages as images, text, audio, frames). It works in the browser and, for what the browser cannot do, in a server container.

Jev

TypeSafe's model that acts as judge: it reranks results, decides whether a passage supports a claim and chooses between page-number candidates.

Reader (lector)

The model that reads pages as images and returns their faithful text, notes, header and footer, and the folio it sees printed.

Manicule (manícula)

The little hand with an outstretched finger that early readers drew in margins to point at a passage. It is Scholaris's emblem.

Pliego

In printing, the large sheet that, folded, makes a gathering of the book. In Scholaris, the group of four pages sent at once to the reader.

Provenance (procedencia)

Where each piece of data comes from: which source gave the record, which reader read the page, how confidently and who corrected it.

Catchword (reclamo)

In early books, the word printed at the foot of a page that repeats the first word of the next. Scholaris keeps it out of the passage text.

SPDF

Scholaris's open format for a document already read: a gzip-compressed SQLite database with text, anchors, figures, vectors and provenance. Current version: 4.1.

.scholaris package

A whole library in a ZIP: one .spdf per document, a manifest and a rights notice.

Stretch (tramo)

In audio and video, the citable unit: 30 to 60 seconds of transcript ending at a sentence boundary.

Unit (unidad)

What is cited as a whole: a page, a stretch of audio, a slide, a chunk of a spreadsheet.

Watcher (vigilante)

A saved search that repeats itself and tells you when what it finds changes.