The tool by Gérard Huet and Pawan Goyal available at https://sanskrit.inria.fr/ from the French Institute for Research in Computer Science and Automation provides a rich collection of tools for analysis of the sanskrit language.
However: it requires an outdated version of ocaml (4.0.7 circa 2018) and many now deprecated features to function.
Because of the challenges involved with running the service this project provides a 'devenv shell' leveraging nixpkgs for reproducing the build environment.
The heritage project architecture depends on httpd running ocaml cgi-scripts over local xml datasets.
httpd installations are requiring non-root users so we will first require a 'heritage' linux* user to run the platform.
- tested on x86 linux but may work in other environments as well
prerequisites:
- server with nix and devenv installed
- a non root user that can successfully run
devenv shell. see: scripts/add-heritage-user.sh
once you have the user we can begin, with a copy of this repository checked out to somewhere like /var/lib/sanskrit-heritage
# assuming you have bash on your system installed:
sudo -u heritage -H bash --login
cd /var/lib/sanskrit-hertiage && devenv shell
# ...
# continue with setup processRepositories making up the heritage platform:
- Zen
- Heritage_Resources
- Heritage_Platform
Steps:
- check all 3 repos out
- create heritage platform config and symlink
- ./configure && make && make install
- run httpd over cgi/htdocs data
After completing above steps you should have a running heritage platform available at http://localhost:4040
Next steps can include using tools like caddy for layering security and authorization.