Data is generated using src/generate.xsl
and served under dist
.
In addition to this, the script also serves to fetch export files from the working environment (parzDB database). As fetching the nearly 900 files takes a while, it is recommended to execute the function only when needed (e.g. after significant data updates).
Lastly, the repository also contains a template for a TEIPublisher application (src/teipb
) serving as a base for a xar
file created during the CI procedures. This file is deployed at https://dhbern.github.io/parzival-static-api/parzival-*.xar
and fetched by the pop-up publisher instance (see https://github.com/DHBern/presentation_parzival for more context).
-
Execute the wrapper script with one of two options:
./parzival-static-api.sh --generate
to generate outputs fordist/api
./parzival-static-api.sh --fetch-exports
to collect export files todist/api/export
-
Alternatively:
- Move to
src
directory - Run
generate.xsl
with an XSLT processor (≥ 3.0).- (Optionally) supply tasks to be executed using the
do
parameter (multiple tasks may be given space separated within quotation marks). - (Optionally) set
verbose=true
to track the execution of the pipeline. - To run all tasks use
all=true
instead ofdo
. - Example command:
java -jar $path-to-saxon/saxon-he-12.5.jar -s:generate.xsl -xsl:generate.xsl do='pass-through-originals contiguous-ranges
.
- (Optionally) supply tasks to be executed using the
- Move to
- Consider CI using Github actions with Node.js and
SaxonJS
. - Consider some degree of "dynamisation" using
SaxonJS
. - Discern sensible and clear-cut border between static backend and TEIPublisher; how far can we push the border?