This directory contains examples of how to use the library from the different languages we support (other than C++. For C++ examples see the tests or tools directories).

At the moment we support C, C++, Rust, Python, and dotnet. The C API has been designed to make it straightforward to wrap for other languages. To see some examples, look at the wrappers directory.

C++

You can begin by looking at example.cc to see some sample usage of the API. custom_builtin.cc shows how to add your own custom builtins (only supported in C++ at this point in time).

From there, the best way to get a more in-depth understanding of the C++ API is to look at the rego tool.

C

There are two C examples. The first, example.c exercises the C API, showing all the various ways it can be used. The second C example is a command line tool that takes one or more Rego data, module, and input files and evaluates a query against them.

To build it on Linux:

a@host:rego-cpp$ cd examples/c
a@host:c$ mkdir build
a@host:c$ cd build
a@host:build$ cmake .. --preset release-clang
a@host:build$ ninja install

and on Windows:

cd examples\c
mkdir build
cd build
cmake .. --preset release
cmake --build . --config Release --target INSTALL

The resulting executable can be called from within the dist directory:

a@host:build$ cd dist
a@host:dist$ ./bin/regoc -d examples/scalars.rego -q data.scalars.greeting
{"expressions":["Hello"]}

Python

The Python example is a simple command line tool that takes zero or more Rego data, module, and input files and evaluates a query against them.

Examples:

a@host:python$ pip install regopy
a@host:python$ python rego.py eval -d examples/scalars.rego data.scalars.greeting
{"expressions":["Hello"]}
 
a@host:python$ python rego.py eval -d examples/objects.rego data.objects.sites[1].name
{"expressions":["smoke1"]}
 
a@host:python$ python rego.py eval -d examples/data0.json -d examples/data1.json -d examples/objects.rego -i examples/input0.json "[data.one, input.b, data.objects.sites[1]]"
{"expressions":[[{"bar":"Foo", "baz":5, "be":true, "bop":23.4},"20",{"name":"smoke1"}]]}
 
a@host:python$ python rego.py eval "x=5; y=x + (2 - 4 * 0.25) * -3 + 7.4;2 * 5"
{"expressions":[true, true, 10], "bindings":{"x":5, "y":9.4}}
 
a@host:python$ python rego.py eval -d examples/bodies.rego -i examples/input1.json data.bodies.e
{"expressions":[{"one":15, "two":15}]}

Rust

The Rust example is another simple command line tool that takes zero or more Rego data, module, and input files and evaluates a query against them.

Examples:

a@host:rust$ cargo run -- eval -d examples/scalars.rego data.scalars.greeting
{"expressions":["Hello"]}
 
a@host:rust$ cargo run -- eval -d examples/objects.rego data.objects.sites[1].name
{"expressions":["smoke1"]}
 
a@host:rust$ cargo run -- eval -d examples/data0.json -d examples/data1.json -d examples/objects.rego -i examples/input0.json "[data.one, input.b, data.objects.sites[1]]"
{"expressions":[[{"bar":"Foo", "baz":5, "be":true, "bop":23.4},"20",{"name":"smoke1"}]]}
 
a@host:rust$ cargo run -- eval "x=5; y=x + (2 - 4 * 0.25) * -3 + 7.4;5 * 2"
{"expressions":[true, true, 10], "bindings":{"x":5, "y":9.4}}
 
a@host:rust$ cargo run -- eval -d examples/bodies.rego -i examples/input1.json data.bodies.e
{"expressions":[{"one":15, "two":15}]}

dotnet

The dotnet example does not provide a full command-line tool, but it does show how use all the features of the wrapper:

a@host:dotnet$ dotnet run
Query Only
{"expressions":[true, true, 10], "bindings":{"x":5, "y":9.4}}
x = 5
10
 
Input and Data
{"expressions":[true], "bindings":{"x":[{"bar":"Foo", "baz":5, "be":true, "bop":23.4},"20",{"name":"smoke1"}]}}
 
Bundles
query: {"expressions":[true], "bindings":{"x":4460}}
example/foo: {"expressions":[2275]}