This is the code for the web interface of SigLA. We describe the organisation of the code as well as the format of the database.
The source code is available here.
The SigLA databases
First, let us describe the SigLA databases. There are two databases:
- One for the documents (tablets)
- One for general metadata about signs (eg. phonetic value, or structure).
The database is a simple directory with: - A
sign.json file containing the database of signs (see below) - A certain hierarchy of directories and subdirectories, whose name do not matter. Each leaf directory represents a document, and should contain the following files: - A file
<document name>.png which should be the image of the whole document. - A file
<document name>_<occurrence number>.png per sign occurrrence on the tablet - A file
meta.json holding information about the database.
The format of
sign.json should be a list of
sign, as described by the OCaml code in
core/sign.mli. The format of
meta.json should be an object adhering to the OCaml type in
The architecture of the code
core/defines alll common data structures of the database shared by all components of sigla-web
client/contains the code of the client, executed in the browser
server/is the code of the websocket server
importis the code of the importer from Krita files.