Introduction

openradtool ("ort") translates a data model (data layout and operations) into C, SQL (SQLite3), and TypeScript code. It also has a raft of related features, such as role-based access control, auditing, full documentation of generated sources, etc.

To keep up to date with the current stable release of openradtool, visit https://kristaps.bsd.lv/openradtool. The website also contains canonical installation, deployment, examples, and usage documentation.

Installation

Clone or update your sources. Configure with ./configure, compile with make (BSD make, so it may be bmake on your system), then make install (or use sudo or doas, if applicable).

To install in an alternative directory to /usr/local, set the PREFIX variable when you run configure.

./configure PREFIX=$HOME/.local
make
make install

If you plan on using pkg-config with the above invocation, make sure that ~/.local/lib/pkgconfig is recognised as a path to package specifications. You'll also want to make sure that man can access the installed location of ~/.local/man, in this case.

Good luck!

Tests

It's useful to run the installed regression tests on the bleeding edge sources. (Again, this uses BSD make, so it may be bmake on your system.) The full regression suite is extensive and may take several minutes to run to completion.

None of the tests require an external network, though some will use an internal network connection to test client/server mode.

make regress

The TypeScript and Node.js regression tests will use Node.js tools if available, though it will not configure these for you. You'll need npm for this to work.

npm install
make regress

This will use the package.json and package-lock.json files to configure the node_modules subdirectory required for the regression tests. You can safely remove the node_modules directory at any time, though you will need to re-run npm install to recreate it. Once installed, the regression suite will automatically pick these up.

To enable picking up the Node.js tests, set the TS_NODE variable to the ts-node script, usually node_modules/.bin/ts-node, in the Makefile. Alternatively, drop it into a Makefile.local that will be included by the Makefile.

To test JSON output, you'll need both the Node.js installation (as described above) and the py-jsonschema (sometimes listed as py3-jsonschema) utilities installed for validating JSON against a schema.

To enable picking up the JSON tests, set the TS_JSONSCHEMA variable to the typescript-json-shema script, usually node_modules/.bin/typescript-json-schema, in the Makefile. Or alternatively, again, in a Makefile.local.

The C interface tests need kcgi, libcurl, sqlbox, and sqlite3 installed. The regression suite will automatically pick these up.

The XLIFF tests need xmllint to validate the XLIFF output. The XSD files are included in the regression suite, so no network connection is made to fetch them.

To enable picking up the XLIFF tests, set the XMLLINT variable to the xmllint program in the Makefile or a Makefile.local.

The rust tests need cargo to operate. Prior to running the tests, dependencies must be downloaded. Regression builds are all performed in offline mode. (The addition of the main.rs file is just to silence the target test. The file will be overwritten when regression tests are run.)

cd rust
mkdir -p rust/src
echo "fn main(){}" > rust/src/main.rs
cargo fetch
cd ..
make regress

As usual, you'll need a CARGO variable in the Makefile or Makefile.local set to the cargo binary.

License

All sources use the ISC (like OpenBSD) license. See the LICENSE.md file for details.