<h1>Rego Treesitter implementation</h1>
For daily routine, I spend majority of the time in the NeoVim, where I write policies
for corporate application. By default, <a href="https://neovim.io/">NeoVim</a> provides the Regex system, which parses
default OPA policies. But I needed more of that, so I started on investigation
how can I achieve a better experience.
<h2>Treesitter libraries</h2>
I don&#39;t need to copy-paste the documentation from the official docs, so to get the 
general principle of it, you can see <a href="https://tree-sitter.github.io/tree-sitter/using-parsers#getting-started">this page</a>.
In short, you will need a repo. For this case, my repo will be <code>tree-sitter-rego</code>.
Initialize the reporisory for the npm
<pre><code class="hljs language-bash">yarn init -y
</code></pre>This will create a package.json with defaults, which you can modify
Then install the necessary dependencies:
<pre><code class="hljs language-bash">yarn add -D tree-sitter-cli nan
</code></pre>Then it&#39;s necessary to create a file for the grammar. Name it <code>grammar.js</code>, which is the convention to name for the tree-sitter.
Fill it with following to create a rego parser:
<pre><code class="hljs language-javascript">module.exports = grammar({
 name: &#x27;rego&#x27;,

 rules: {
 source_file: $ =&gt; repeat($._definition),

 _definition: $ =&gt; choice(
 $.package_definition,
 $.import_package,
 $.operator_check,
 $.comment,
 $.rego_block,
 $.builtin_function,
 $._junk
 ),

 operator: $ =&gt; choice(
 &#x27;==&#x27;,
 &#x27;:=&#x27;,
 &#x27;=&#x27;,
 &#x27;!=&#x27;,
 &#x27;&lt;&#x27;,
 &#x27;&gt;&#x27;,
 &#x27;/&#x27;,
 &#x27;-&#x27;,
 &#x27;+&#x27;,
 ),

 true: $ =&gt; &#x27;true&#x27;,
 false: $ =&gt; &#x27;false&#x27;,

 comma: $ =&gt; &#x27;,&#x27;,
 
 comment: $ =&gt; /\#.*?\n\r?/,

 function_name: $ =&gt; choice(
 &#x27;lower&#x27;,
 &#x27;is_string&#x27;,
 &#x27;object.get&#x27;,
 &#x27;print&#x27;,
 &#x27;concat&#x27;,
 &#x27;contains&#x27;,
 &#x27;time.now_ns&#x27;,
 &#x27;io.jwt.encode_sign_raw&#x27;,
 &#x27;io.jwt.encode_sign&#x27;,
 &#x27;io.jwt.decode&#x27;,
 &#x27;io.jwt.verify_es256&#x27;,
 &#x27;strings.replace_n&#x27;,
 &#x27;http.send&#x27;,
 ),

 opening_parameter: $ =&gt; &#x27;(&#x27;,

 closing_parameter: $ =&gt; &#x27;)&#x27;,

 builtin_function: $ =&gt; seq(
 field(&#x27;function_name&#x27;, 
 $.function_name
 ),
 field(&#x27;opening_parameter&#x27;, $.opening_parameter),
 field(&#x27;function_body&#x27;, 
 repeat(
 choice(
 $.identifier,
 $.array_definition,
 $.true,
 $.false,
 $.number,
 $.object_field,
 $.string_definition,
 $.identifier,
 $.comma,
 ),
 ),
 ),
 field(&#x27;closing_parameter&#x27;, $.closing_parameter),
 ),

 string_definition: $ =&gt; seq(
 &#x27;&quot;&#x27;,
 /[a-zA-Z0-9&lt;&gt;@\-._:=\s\/\\]*/,
 &#x27;&quot;&#x27;,
 ),

 _array_opening: $ =&gt; &#x27;[&#x27;,
 _array_closing: $ =&gt; &#x27;]&#x27;,

 object_field: $ =&gt; prec(
 1, 
 seq(
 /[a-zA-Z\._]+\[/,
 choice(
 $.identifier,
 $.number,
 $.object_field,
 $.string_definition
 ),
 $._array_closing,
 ),
 ),

 array_definition: $ =&gt; seq(
 $._array_opening,
 repeat(
 choice(
 $.array_definition,
 $.string_definition,
 $.identifier,
 $.identifier,
 $.number,
 $.object_field,
 $.true,
 $.false,
 $.comma,
 ),
 ),
 $._array_closing,
 ),

 operator_check: $ =&gt; seq(
 choice(
 $.identifier,
 $.builtin_function,
 $.string_definition,
 $.object_field,
 $.array_definition,
 $.true,
 $.false,
 ),
 $.operator,
 choice(
 $.identifier,
 $.builtin_function,
 $.string_definition,
 $.object_field,
 $.array_definition,
 $.true,
 $.false,
 ),
 ),

 rego_rule: $ =&gt; prec(1, choice(
 $.identifier,
 $.operator_check,
 $.array_definition,
 $.test_case,
 $.true,
 $.false,
 ),),

 test_case: $ =&gt; seq(
 $.identifier,
 repeat(
 seq(
 $.reserved_keywords,
 $.identifier,
 ),
 ),
 ),

 rego_block: $ =&gt; seq(
 field(&#x27;rego_rule_name&#x27;, $.identifier),
 optional(
 seq(
 $.operator,
 $.identifier,
 ),
 ),
 &#x27;{&#x27;,
 repeat($.rego_rule),
 &#x27;}&#x27;,
 ),

 _junk: $ =&gt; /\n/,

 reserved_keywords: $ =&gt; choice(
 &#x27;as&#x27;,
 &#x27;with&#x27;,
 ),
 
 as_keyword: $ =&gt; seq(
 $.reserved_keywords,
 field(&#x27;package_alias&#x27;, $.identifier),
 ),

 import_package: $ =&gt; seq(
 &#x27;import&#x27;,
 field(&#x27;imported_package_name&#x27;, 
 choice(
 $.identifier,
 ),
 ),
 optional($.as_keyword),
 ),

 package_definition: $ =&gt; seq(
 &#x27;package&#x27;,
 field(&#x27;package_name&#x27;, $.identifier),
 ),

 identifier: $ =&gt; /[a-zA-Z\._]+/,

 number: $ =&gt; /\d+/
 }
});
</code></pre>After this, you can execute a command <code>tree-sitter generate</code>. For convenience, you can even set up
the aliases in scripts for package.json, so you will be able to verify the correctness of your 
grammar.
<pre><code class="hljs language-json">{
 &quot;scripts&quot;: {
 &quot;generate&quot;: &quot;tree-sitter generate&quot;,
 &quot;build&quot;: &quot;tree-sitter generate &amp;&amp; node-gyp configure &amp;&amp; node-gyp build&quot;,
 &quot;test&quot;: &quot;tree-sitter test&quot;
 }
}
</code></pre>Complete source for the parser located here: <a href="https://github.com/FallenAngel97/tree-sitter-rego">https://github.com/FallenAngel97/tree-sitter-rego</a>
Photo by <a href="https://unsplash.com/@tanelah?utm_source=unsplash&utm_medium=referral&utm_content=creditCopyText">Taneli Lahtinen</a> on <a href="https://unsplash.com/?utm_source=unsplash&utm_medium=referral&utm_content=creditCopyText">Unsplash</a>