minimal working grammar

README.md

@@ -10,9 +10,13 @@ non-flake equivalents:
 - `nix-shell`
 
 
-99% of this repository is code-gen'd garbage.
-to recreate it or update it, follow this guide:
+## development workflow
+guide on using tree-sitter:
 <https://tree-sitter.github.io/tree-sitter/creating-parsers>
 
-recording my steps here to maybe someday make this reproducible:
+roughly:
+1. `nix develop`
+2. `tree-sitter generate`
+3. `tree-sitter test`
 
+the parser is defined in `grammar.js`, and tests live in `corpus/`.

corpus/shebang_only.txt

@@ -0,0 +1,14 @@
+==================
+nix-shell shebang without any interpreter option
+==================
+
+#!/usr/bin/env nix-shell
+---
+
+(source_file
+  (first_line
+    (shebang_open)
+    (opt_ws)
+    (env)
+    (ws)
+    (nix_shell)))

corpus/shebang_with_sub_interpreter.txt

@@ -0,0 +1,22 @@
+==================
+Minimal nix-shell invocation
+==================
+
+#!/usr/bin/env nix-shell
+#!nix-shell -i bash
+---
+
+(source_file
+  (first_line
+    (shebang_open)
+    (opt_ws)
+    (env)
+    (ws)
+    (nix_shell))
+  (non_first_line
+    (newline)
+    (annotation_line
+      (shebang_open)
+      (opt_ws)
+      (nix_shell)
+      (nix_shell_args))))

grammar.js

@@ -2,7 +2,20 @@ module.exports = grammar({
   name: 'nixshell',
 
   rules: {
-    // TODO: add the actual grammar rules
-    source_file: $ => 'hello'
+    source_file: $ => seq($.first_line, repeat($.non_first_line)),
+
+    first_line: $ => seq($.shebang_open, $.opt_ws, optional(seq($.env, $.ws)), $.nix_shell),
+    non_first_line: $ => seq($.newline, $.annotation_line),
+    annotation_line: $ => seq($.shebang_open, $.opt_ws, $.nix_shell, $.nix_shell_args),
+    embedded_script_line: $ => /.*/,
+
+    shebang_open: $ => /#!/,
+    // TODO: allow `usr/bin/env`, `env`
+    env: $ => '/usr/bin/env',
+    nix_shell: $ => 'nix-shell',
+    nix_shell_args: $ => ' -i bash',
+    ws: $ => / +/,
+    opt_ws: $ => / */,
+    newline: $ => '\n',
   }
 });