Initial working

.gitignore

@@ -0,0 +1,6 @@
+Cargo.lock
+node_modules
+build
+*.log
+package-lock.json
+target

Cargo.toml

@@ -0,0 +1,26 @@
+[package]
+name = "tree-sitter-dockerfile"
+description = "dockerfile grammar for the tree-sitter parsing library"
+version = "0.0.1"
+keywords = ["incremental", "parsing", "dockerfile"]
+categories = ["parsing", "text-editors"]
+repository = "https://github.com/tree-sitter/tree-sitter-javascript"
+edition = "2018"
+license = "MIT"
+
+build = "bindings/rust/build.rs"
+include = [
+  "bindings/rust/*",
+  "grammar.js",
+  "queries/*",
+  "src/*",
+]
+
+[lib]
+path = "bindings/rust/lib.rs"
+
+[dependencies]
+tree-sitter = "0.17"
+
+[build-dependencies]
+cc = "1.0"

binding.gyp

@@ -0,0 +1,19 @@
+{
+  "targets": [
+    {
+      "target_name": "tree_sitter_dockerfile_binding",
+      "include_dirs": [
+        "<!(node -e \"require('nan')\")",
+        "src"
+      ],
+      "sources": [
+        "bindings/node/binding.cc",
+        "src/parser.c",
+        # If your language uses an external scanner, add it here.
+      ],
+      "cflags_c": [
+        "-std=c99",
+      ]
+    }
+  ]
+}

bindings/node/binding.cc

@@ -0,0 +1,28 @@
+#include "tree_sitter/parser.h"
+#include <node.h>
+#include "nan.h"
+
+using namespace v8;
+
+extern "C" TSLanguage * tree_sitter_dockerfile();
+
+namespace {
+
+NAN_METHOD(New) {}
+
+void Init(Local<Object> exports, Local<Object> module) {
+  Local<FunctionTemplate> tpl = Nan::New<FunctionTemplate>(New);
+  tpl->SetClassName(Nan::New("Language").ToLocalChecked());
+  tpl->InstanceTemplate()->SetInternalFieldCount(1);
+
+  Local<Function> constructor = Nan::GetFunction(tpl).ToLocalChecked();
+  Local<Object> instance = constructor->NewInstance(Nan::GetCurrentContext()).ToLocalChecked();
+  Nan::SetInternalFieldPointer(instance, 0, tree_sitter_dockerfile());
+
+  Nan::Set(instance, Nan::New("name").ToLocalChecked(), Nan::New("dockerfile").ToLocalChecked());
+  Nan::Set(module, Nan::New("exports").ToLocalChecked(), instance);
+}
+
+NODE_MODULE(tree_sitter_dockerfile_binding, Init)
+
+}  // namespace

bindings/node/index.js

@@ -0,0 +1,19 @@
+try {
+  module.exports = require("../../build/Release/tree_sitter_dockerfile_binding");
+} catch (error1) {
+  if (error1.code !== 'MODULE_NOT_FOUND') {
+    throw error1;
+  }
+  try {
+    module.exports = require("../../build/Debug/tree_sitter_dockerfile_binding");
+  } catch (error2) {
+    if (error2.code !== 'MODULE_NOT_FOUND') {
+      throw error2;
+    }
+    throw error1
+  }
+}
+
+try {
+  module.exports.nodeTypeInfo = require("../../src/node-types.json");
+} catch (_) {}

bindings/rust/build.rs

@@ -0,0 +1,40 @@
+fn main() {
+    let src_dir = std::path::Path::new("src");
+
+    let mut c_config = cc::Build::new();
+    c_config.include(&src_dir);
+    c_config
+        .flag_if_supported("-Wno-unused-parameter")
+        .flag_if_supported("-Wno-unused-but-set-variable")
+        .flag_if_supported("-Wno-trigraphs");
+    let parser_path = src_dir.join("parser.c");
+    c_config.file(&parser_path);
+
+    // If your language uses an external scanner written in C,
+    // then include this block of code:
+
+    /*
+    let scanner_path = src_dir.join("scanner.c");
+    c_config.file(&scanner_path);
+    println!("cargo:rerun-if-changed={}", scanner_path.to_str().unwrap());
+    */
+
+    c_config.compile("parser");
+    println!("cargo:rerun-if-changed={}", parser_path.to_str().unwrap());
+
+    // If your language uses an external scanner written in C++,
+    // then include this block of code:
+
+    /*
+    let mut cpp_config = cc::Build::new();
+    cpp_config.cpp(true);
+    cpp_config.include(&src_dir);
+    cpp_config
+        .flag_if_supported("-Wno-unused-parameter")
+        .flag_if_supported("-Wno-unused-but-set-variable");
+    let scanner_path = src_dir.join("scanner.cc");
+    cpp_config.file(&scanner_path);
+    cpp_config.compile("scanner");
+    println!("cargo:rerun-if-changed={}", scanner_path.to_str().unwrap());
+    */
+}

bindings/rust/lib.rs

@@ -0,0 +1,52 @@
+//! This crate provides dockerfile language support for the [tree-sitter][] parsing library.
+//!
+//! Typically, you will use the [language][language func] function to add this language to a
+//! tree-sitter [Parser][], and then use the parser to parse some code:
+//!
+//! ```
+//! let code = "";
+//! let mut parser = tree_sitter::Parser::new();
+//! parser.set_language(tree_sitter_dockerfile::language()).expect("Error loading dockerfile grammar");
+//! let tree = parser.parse(code, None).unwrap();
+//! ```
+//!
+//! [Language]: https://docs.rs/tree-sitter/*/tree_sitter/struct.Language.html
+//! [language func]: fn.language.html
+//! [Parser]: https://docs.rs/tree-sitter/*/tree_sitter/struct.Parser.html
+//! [tree-sitter]: https://tree-sitter.github.io/
+
+use tree_sitter::Language;
+
+extern "C" {
+    fn tree_sitter_dockerfile() -> Language;
+}
+
+/// Get the tree-sitter [Language][] for this grammar.
+///
+/// [Language]: https://docs.rs/tree-sitter/*/tree_sitter/struct.Language.html
+pub fn language() -> Language {
+    unsafe { tree_sitter_dockerfile() }
+}
+
+/// The content of the [`node-types.json`][] file for this grammar.
+///
+/// [`node-types.json`]: https://tree-sitter.github.io/tree-sitter/using-parsers#static-node-types
+pub const NODE_TYPES: &'static str = include_str!("../../src/node-types.json");
+
+// Uncomment these to include any queries that this grammar contains
+
+// pub const HIGHLIGHTS_QUERY: &'static str = include_str!("../../queries/highlights.scm");
+// pub const INJECTIONS_QUERY: &'static str = include_str!("../../queries/injections.scm");
+// pub const LOCALS_QUERY: &'static str = include_str!("../../queries/locals.scm");
+// pub const TAGS_QUERY: &'static str = include_str!("../../queries/tags.scm");
+
+#[cfg(test)]
+mod tests {
+    #[test]
+    fn test_can_load_grammar() {
+        let mut parser = tree_sitter::Parser::new();
+        parser
+            .set_language(super::language())
+            .expect("Error loading dockerfile language");
+    }
+}

corpus/add.txt

@@ -0,0 +1,27 @@
+==================
+No param
+==================
+
+ADD /src ./dst
+
+---
+
+(source_file
+	(add_instruction
+		(path)
+		(path)))
+
+==================
+With param
+==================
+
+ADD --chown=a:b /src ./dst
+
+---
+
+(source_file
+	(add_instruction
+		(param)
+		(path)
+		(path)))
+

corpus/arg.txt

@@ -0,0 +1,40 @@
+==================
+No default
+==================
+
+ARG test
+
+---
+
+(source_file
+	(arg_instruction
+		name: (unquoted_string)))
+
+==================
+With default
+==================
+
+ARG foo=bar
+
+---
+
+(source_file
+	(arg_instruction
+		name: (unquoted_string)
+		default: (unquoted_string)))
+		
+
+==================
+With quoted default
+==================
+
+ARG foo="bar"
+
+---
+
+(source_file
+	(arg_instruction
+		name: (unquoted_string)
+		default: (double_quoted_string)))
+		
+

corpus/cmd.txt

@@ -0,0 +1,53 @@
+==================
+Shell command
+==================
+
+CMD echo "test"
+
+---
+
+(source_file
+	(cmd_instruction
+		(shell_command)))
+
+==================
+Shell command multiline
+==================
+
+CMD echo "test" \
+	"foo" \
+	bar
+
+
+---
+
+(source_file
+	(cmd_instruction
+		(shell_command)))
+
+==================
+Run with shell empty array
+==================
+
+cmd []
+
+---
+
+(source_file
+	(cmd_instruction
+	    (string_array)))
+
+==================
+Run with shell array
+==================
+
+cmd ["echo", "test"]
+
+---
+
+(source_file
+	(cmd_instruction
+	    (string_array
+		    (double_quoted_string)
+			(double_quoted_string))))
+

corpus/copy.txt

@@ -0,0 +1,27 @@
+==================
+No param
+==================
+
+COPY /src ./dst
+
+---
+
+(source_file
+	(copy_instruction
+		(path)
+		(path)))
+
+==================
+With param
+==================
+
+COPY --chown=a:b /src ./dst
+
+---
+
+(source_file
+	(copy_instruction
+		(param)
+		(path)
+		(path)))
+

corpus/entrypoint.txt

@@ -0,0 +1,53 @@
+==================
+Shell command
+==================
+
+ENTRYPOINT echo "test"
+
+---
+
+(source_file
+	(entrypoint_instruction
+		(shell_command)))
+
+==================
+Shell command multiline
+==================
+
+ENTRYPOINT echo "test" \
+	"foo" \
+	bar
+
+
+---
+
+(source_file
+	(entrypoint_instruction
+		(shell_command)))
+
+==================
+Run with shell empty array
+==================
+
+ENTRYPOINT []
+
+---
+
+(source_file
+	(entrypoint_instruction
+	    (string_array)))
+
+==================
+Run with shell array
+==================
+
+ENTRYPOINT ["echo", "test"]
+
+---
+
+(source_file
+	(entrypoint_instruction
+	    (string_array
+		    (double_quoted_string)
+			(double_quoted_string))))
+

corpus/env.txt

@@ -0,0 +1,81 @@
+==================
+Quoted value
+==================
+
+ENV TEST="okay"
+
+---
+
+(source_file
+	(env_instruction
+		(env_pair
+			name: (unquoted_string)
+			value: (double_quoted_string))))
+
+==================
+Unquoted value
+==================
+
+ENV TEST_2=value\ 2
+
+---
+
+(source_file
+	(env_instruction
+		(env_pair
+			name: (unquoted_string)
+			value: (unquoted_string))))
+
+==================
+Multiple
+==================
+
+ENV TEST="foo" TEST_2=foo\ bar
+
+---
+
+(source_file
+	(env_instruction
+		(env_pair
+			name: (unquoted_string)
+			value: (double_quoted_string))
+		(env_pair
+			name: (unquoted_string)
+			value: (unquoted_string))))
+
+==================
+Multiline
+==================
+
+ENV TEST1="foo" \
+	TEST2=bar
+
+---
+
+(source_file
+	(env_instruction
+		(env_pair
+			name: (unquoted_string)
+			value: (double_quoted_string))
+		(env_pair
+			name: (unquoted_string)
+			value: (unquoted_string))))
+
+==================
+Multiple instructions
+==================
+
+ENV TEST1="foo" 
+ENV TEST2="bar" 
+
+---
+
+(source_file
+	(env_instruction
+		(env_pair
+			name: (unquoted_string)
+			value: (double_quoted_string)))
+	(env_instruction
+		(env_pair
+			name: (unquoted_string)
+			value: (double_quoted_string))))

corpus/expose.txt

@@ -0,0 +1,27 @@
+==================
+Single port
+==================
+
+EXPOSE 80
+
+
+---
+
+(source_file
+	(expose_instruction
+		(expose_port)))
+
+==================
+Multiple ports
+==================
+
+EXPOSE 80 90/tcp 100/udp
+
+
+---
+
+(source_file
+	(expose_instruction
+		(expose_port)
+		(expose_port)
+		(expose_port)))

corpus/from.txt

@@ -0,0 +1,68 @@
+==================
+Only image
+==================
+
+FROM testimage
+
+---
+
+(source_file
+	(from_instruction
+		(image_spec)))
+
+==================
+Image with tag
+==================
+
+FROM testimage:tag
+
+---
+
+(source_file
+	(from_instruction
+		(image_spec
+		    (image_tag))))
+
+==================
+Image with digest
+==================
+
+FROM testimage@sha256:skgshlshg
+
+---
+
+(source_file
+	(from_instruction
+		(image_spec
+		    (image_digest))))
+
+==================
+Image with tag and digest
+==================
+
+FROM testimage:tag@sha256:452525
+
+---
+
+(source_file
+	(from_instruction
+		(image_spec
+		    (image_tag)
+		    (image_digest))))
+
+
+==================
+Tag and digest with as
+==================
+
+FROM sourcegraph/alpine:3.12@sha256:ce099fbcd3cf70b338fc4cb2a4e1fa9ae847de21afdb0a849a393b87d94fb174 AS monitoring_builder
+
+---
+
+(source_file
+	(from_instruction
+		(image_spec
+		    (image_tag)
+		    (image_digest))
+		(name)))
+

corpus/healthcheck.txt

@@ -0,0 +1,39 @@
+==================
+None
+==================
+
+HEALTHCHECK NONE
+
+---
+
+(source_file
+	(healthcheck_instruction))
+
+==================
+Custom command
+==================
+
+HEALTHCHECK CMD	echo "test"
+
+---
+
+(source_file
+	(healthcheck_instruction
+		(cmd_instruction
+			(shell_command))))
+
+==================
+With options
+==================
+
+HEALTHCHECK --interval=10s --retries=10 CMD echo "test"
+
+---
+
+(source_file
+	(healthcheck_instruction
+		(param)
+		(param)
+		(cmd_instruction
+			(shell_command))))
+

corpus/label.txt

@@ -0,0 +1,48 @@
+==================
+Single key-value
+==================
+
+LABEL key="value"
+
+---
+
+(source_file
+	(label_instruction
+		(label_pair
+			key: (unquoted_string)
+			value: (double_quoted_string))))
+
+==================
+Multi key-value
+==================
+
+LABEL key.1="value1" key.2="value2"
+
+---
+
+(source_file
+	(label_instruction
+		(label_pair
+			key: (unquoted_string)
+			value: (double_quoted_string))
+		(label_pair
+			key: (unquoted_string)
+			value: (double_quoted_string))))
+
+==================
+Multiline
+==================
+
+LABEL key.1="value1" \
+	  key.2="value2"
+
+---
+
+(source_file
+	(label_instruction
+		(label_pair
+			key: (unquoted_string)
+			value: (double_quoted_string))
+		(label_pair
+			key: (unquoted_string)
+			value: (double_quoted_string))))

corpus/onbuild.txt

@@ -0,0 +1,26 @@
+==================
+Simple add
+==================
+
+ONBUILD ADD . /app/src
+
+---
+
+(source_file
+	(onbuild_instruction
+		(add_instruction
+			(path)
+			(path))))
+
+==================
+Simple run
+==================
+
+ONBUILD RUN /usr/local/bin/python-build --dir /app/src
+
+---
+
+(source_file
+	(onbuild_instruction
+		(run_instruction
+			(shell_command))))

corpus/run.txt

@@ -0,0 +1,53 @@
+==================
+Shell command
+==================
+
+RUN echo "test"
+
+---
+
+(source_file
+	(run_instruction
+		(shell_command)))
+
+==================
+Shell command multiline
+==================
+
+RUN echo "test" \
+	"foo" \
+	bar
+
+
+---
+
+(source_file
+	(run_instruction
+		(shell_command)))
+
+==================
+Run with shell empty array
+==================
+
+run []
+
+---
+
+(source_file
+	(run_instruction
+	    (string_array)))
+
+==================
+Run with shell array
+==================
+
+run ["echo", "test"]
+
+---
+
+(source_file
+	(run_instruction
+	    (string_array
+		    (double_quoted_string)
+			(double_quoted_string))))
+

corpus/shell_instruction.txt

@@ -0,0 +1,14 @@
+==================
+Simple
+==================
+
+SHELL ["powershell","-command"]
+
+---
+
+(source_file
+	(shell_instruction
+		(string_array
+			(double_quoted_string)
+			(double_quoted_string))))
+

corpus/stopsignal.txt

@@ -0,0 +1,21 @@
+==================
+Integer
+==================
+
+STOPSIGNAL 9
+
+---
+
+(source_file
+	(stopsignal_instruction))
+
+==================
+Signal name
+==================
+
+STOPSIGNAL SIGKILL
+
+---
+
+(source_file
+	(stopsignal_instruction))

corpus/user.txt

@@ -0,0 +1,26 @@
+==================
+No group
+==================
+
+USER foo
+
+---
+
+(source_file
+	(user_instruction
+		user: (unquoted_string)))
+
+==================
+With group
+==================
+
+USER foo:bar
+
+---
+
+(source_file
+	(user_instruction
+		user: (unquoted_string)
+		group: (unquoted_string)))
+		
+

corpus/volume.txt

@@ -0,0 +1,67 @@
+==================
+Single volume
+==================
+
+VOLUME /myvol
+
+---
+
+(source_file
+	(volume_instruction
+		(path)))
+
+==================
+Multiple volumes
+==================
+
+volume /myvol1 /myvol2
+
+
+---
+
+(source_file
+	(volume_instruction
+		(path)
+		(path)))
+
+==================
+Multiline volumes
+==================
+
+VOLUME /myvol1 \
+	/myvol2
+
+---
+
+(source_file
+	(volume_instruction
+		(path)
+		(path)))
+
+==================
+Array
+==================
+
+VOLUME ["/test/myvol"]
+
+---
+
+(source_file
+	(volume_instruction
+	    (string_array
+			(double_quoted_string))))
+
+==================
+Multiline array
+==================
+
+VOLUME ["/test/myvol", \
+		"/test/myvol2"]
+
+---
+
+(source_file
+	(volume_instruction
+	    (string_array
+			(double_quoted_string)
+			(double_quoted_string))))

corpus/workdir.txt

@@ -0,0 +1,12 @@
+==================
+Simple
+==================
+
+WORKDIR /tmp/test
+
+---
+
+(source_file
+	(workdir_instruction
+		(path)))
+

grammar.js

@@ -0,0 +1,268 @@
+module.exports = grammar({
+		name: 'dockerfile',
+
+		extras: $ => [$.comment, /\s+/, '\\\n'],
+
+		rules: {
+				source_file: $ => repeat(seq($._instruction, "\n")),
+
+				_instruction: $ => choice(
+						$.from_instruction,
+						$.run_instruction,
+						$.cmd_instruction,
+						$.label_instruction,
+						$.expose_instruction,
+						$.env_instruction,
+						$.add_instruction,
+						$.copy_instruction,
+						$.entrypoint_instruction,
+						$.volume_instruction,
+						$.user_instruction,
+						$.workdir_instruction,
+						$.arg_instruction,
+						$.onbuild_instruction,
+						$.stopsignal_instruction,
+						$.healthcheck_instruction,
+						$.shell_instruction,
+				),
+
+				from_instruction: $ => seq(
+						alias(/[fF][rR][oO][mM]/, "FROM"),
+						$._non_newline_whitespace,
+						optional($.param),
+						$.image_spec,
+						optional(seq(
+								alias(/[aA][sS]/, "AS"),
+								$.name,
+						)),
+				),
+
+				run_instruction: $ => seq(
+						alias(/[rR][uU][nN]/, "RUN"),
+						$._non_newline_whitespace,
+						choice(
+								$.string_array,
+								$.shell_command,
+						),
+				),
+
+				cmd_instruction: $ => seq(
+						alias(/[cC][mM][dD]/, "CMD"),
+						$._non_newline_whitespace,
+						choice(
+								$.string_array,
+								$.shell_command,
+						),
+				),
+
+				label_instruction: $ => seq(
+						alias(/[lL][aA][bB][eE][lL]/, "LABEL"),
+						$._non_newline_whitespace,
+						repeat1($.label_pair),
+				),
+
+				expose_instruction: $ => seq(
+						alias(/[eE][xX][pP][oO][sS][eE]/, "EXPOSE"),
+						$._non_newline_whitespace,
+						repeat1($.expose_port),
+				),
+
+				env_instruction: $ => seq(
+						alias(/[eE][nN][vV]/, "ENV"),
+						$._non_newline_whitespace,
+						repeat1($.env_pair),
+				),
+
+				add_instruction: $ => seq(
+						alias(/[aA][dD][dD]/, "ADD"),
+						$._non_newline_whitespace,
+						optional($.param),
+						$.path,
+						$.path,
+				),
+
+				copy_instruction: $ => seq(
+						alias(/[cC][oO][pP][yY]/, "COPY"),
+						$._non_newline_whitespace,
+						optional($.param),
+						$.path,
+						$.path,
+				),
+
+				entrypoint_instruction: $ => seq(
+						alias(/[eE][nN][tT][rR][yY][pP][oO][iI][nN][tT]/, "ENTRYPOINT"),
+						$._non_newline_whitespace,
+						choice(
+								$.string_array,
+								$.shell_command,
+						),
+				),
+
+				volume_instruction: $ => seq(
+						alias(/[vV][oO][lL][uU][mM][eE]/, "VOLUME"),
+						$._non_newline_whitespace,
+						choice(
+								$.string_array,
+								repeat1($.path),
+						),
+				),
+
+				user_instruction: $ => seq(
+						alias(/[uU][sS][eE][rR]/, "USER"),
+						$._non_newline_whitespace,
+						field("user", alias(/[a-z][-a-z0-9_]*/, $.unquoted_string)),
+						optional(seq(
+							token.immediate(":"),
+							field("group", alias(token.immediate(/[a-z][-a-z0-9_]*/), $.unquoted_string)),
+						)),
+				),
+
+				workdir_instruction: $ => seq(
+						alias(/[wW][oO][rR][kK][dD][iI][rR]/, "WORKDIR"),
+						$._non_newline_whitespace,
+						$.path,
+				),
+
+				arg_instruction: $ => seq(
+						alias(/[aA][rR][gG]/, "ARG"),
+						$._non_newline_whitespace,
+						field("name", alias(/[a-zA-Z0-9_]+/, $.unquoted_string)),
+						optional(seq(
+							token.immediate("="),
+							field("default", choice(
+								$.double_quoted_string,
+								$.unquoted_string,
+							)),
+						)),
+				),
+
+				onbuild_instruction: $ => seq(
+						alias(/[oO][nN][bB][uU][iI][lL][dD]/, "ONBUILD"),
+						$._non_newline_whitespace,
+						$._instruction,
+				),
+
+				stopsignal_instruction: $ => seq(
+						alias(/[sS][tT][oO][pP][sS][iI][gG][nN][aA][lL]/, "STOPSIGNAL"),
+						$._non_newline_whitespace,
+						/[A-Z0-9]+/,
+				),
+
+				healthcheck_instruction: $ => seq(
+						alias(/[hH][eE][aA][lL][tT][hH][cC][hH][eE][cC][kK]/, "HEALTHCHECK"),
+						$._non_newline_whitespace,
+						choice(
+							"NONE",
+							seq(
+								repeat($.param),
+								$.cmd_instruction,
+							),
+						),
+				),
+
+				shell_instruction: $ => seq(
+						alias(/[sS][hH][eE][lL][lL]/, "SHELL"),
+						$._non_newline_whitespace,
+						$.string_array,
+				),
+				
+				path: $ => /[^-\[][^\s]*/,
+
+				env_pair: $ => seq(
+					field("name", alias(/[a-zA-Z][a-zA-Z0-9_]+[a-zA-Z0-9]/, $.unquoted_string)),
+					token.immediate("="),
+					field("value", choice(
+						$.double_quoted_string,
+						$.unquoted_string,
+					)),
+				),
+
+				expose_port: $ => seq(
+					/\d+/,
+					optional(choice(
+						"/tcp",
+						"/udp",
+					)),
+				),
+
+				label_pair: $ => seq(
+					field("key", alias(/[-a-zA-Z0-9\._]+/, $.unquoted_string)),
+					token.immediate("="),
+					field("value", choice(
+						$.double_quoted_string,
+						$.unquoted_string,
+					)),
+				),
+
+				image_spec: $ => seq(
+						alias(/[^:@\s]+/, "name"),
+						seq(
+								optional($.image_tag),
+								optional($.image_digest),
+						),
+				),
+
+				image_tag: $ => token.immediate(seq(
+						":",
+						/[^\s@]+/,
+				)),
+
+				image_digest: $ => token.immediate(seq(
+						"@",
+						/[^\s]+/,
+				)),
+
+				param: $ => seq(
+						"--",
+						field("name", /[a-z][-a-z]*/),
+						"=",
+						field("value", /[^\s]+/),
+				),
+
+				name: $ => /[-a-z_]+/,
+
+				string_array: $ => seq(
+						"[",
+						optional(seq(
+								$.double_quoted_string,
+								repeat(seq(",", $.double_quoted_string)),
+						)),
+						"]",
+				),
+
+				shell_command: $ => seq(
+						/[^\[\s].*[^\\\n]/,
+						repeat(seq("\\\n", /.*[^\\\n]/)),
+				),
+
+				double_quoted_string: $ => seq(
+						'"',
+						repeat(choice(
+								token.immediate(prec(1, /[^"\n\\]+/)),
+								$.escape_sequence
+						)),
+						'"'
+				),
+
+				unquoted_string: $ => repeat1(choice(
+					token.immediate(/[^\s\n\"\\]+/),
+					token.immediate("\\ "),
+				)),
+
+				escape_sequence: $ => token.immediate(seq(
+						'\\',
+						choice(
+								/[^xuU]/,
+								/\d{2,3}/,
+								/x[0-9a-fA-F]{2,}/,
+								/u[0-9a-fA-F]{4}/,
+								/U[0-9a-fA-F]{8}/
+						)
+				)),
+
+				_non_newline_whitespace: $ => /[\t ]*/,
+
+				comment: $ => seq("#", /.*/),
+		}
+});
+

package.json

@@ -0,0 +1,17 @@
+{
+  "name": "tree-sitter-dockerfile",
+  "version": "0.1.0",
+  "description": "A tree-sitter module for the Dockerfile grammar",
+  "main": "bindings/node",
+  "scripts": {
+    "test": "echo \"Error: no test specified\" && exit 1"
+  },
+  "author": "Camden Cheek <camden@ccheek.com>",
+  "license": "ISC",
+  "dependencies": {
+    "nan": "^2.14.2"
+  },
+  "devDependencies": {
+    "tree-sitter-cli": "^0.19.4"
+  }
+}

src/node-types.json

@@ -0,0 +1,738 @@
+[
+  {
+    "type": "add_instruction",
+    "named": true,
+    "fields": {},
+    "children": {
+      "multiple": true,
+      "required": true,
+      "types": [
+        {
+          "type": "param",
+          "named": true
+        },
+        {
+          "type": "path",
+          "named": true
+        }
+      ]
+    }
+  },
+  {
+    "type": "arg_instruction",
+    "named": true,
+    "fields": {
+      "default": {
+        "multiple": false,
+        "required": false,
+        "types": [
+          {
+            "type": "double_quoted_string",
+            "named": true
+          },
+          {
+            "type": "unquoted_string",
+            "named": true
+          }
+        ]
+      },
+      "name": {
+        "multiple": false,
+        "required": true,
+        "types": [
+          {
+            "type": "unquoted_string",
+            "named": true
+          }
+        ]
+      }
+    }
+  },
+  {
+    "type": "cmd_instruction",
+    "named": true,
+    "fields": {},
+    "children": {
+      "multiple": false,
+      "required": true,
+      "types": [
+        {
+          "type": "shell_command",
+          "named": true
+        },
+        {
+          "type": "string_array",
+          "named": true
+        }
+      ]
+    }
+  },
+  {
+    "type": "comment",
+    "named": true,
+    "fields": {}
+  },
+  {
+    "type": "copy_instruction",
+    "named": true,
+    "fields": {},
+    "children": {
+      "multiple": true,
+      "required": true,
+      "types": [
+        {
+          "type": "param",
+          "named": true
+        },
+        {
+          "type": "path",
+          "named": true
+        }
+      ]
+    }
+  },
+  {
+    "type": "double_quoted_string",
+    "named": true,
+    "fields": {},
+    "children": {
+      "multiple": true,
+      "required": false,
+      "types": [
+        {
+          "type": "escape_sequence",
+          "named": true
+        }
+      ]
+    }
+  },
+  {
+    "type": "entrypoint_instruction",
+    "named": true,
+    "fields": {},
+    "children": {
+      "multiple": false,
+      "required": true,
+      "types": [
+        {
+          "type": "shell_command",
+          "named": true
+        },
+        {
+          "type": "string_array",
+          "named": true
+        }
+      ]
+    }
+  },
+  {
+    "type": "env_instruction",
+    "named": true,
+    "fields": {},
+    "children": {
+      "multiple": true,
+      "required": true,
+      "types": [
+        {
+          "type": "env_pair",
+          "named": true
+        }
+      ]
+    }
+  },
+  {
+    "type": "env_pair",
+    "named": true,
+    "fields": {
+      "name": {
+        "multiple": false,
+        "required": true,
+        "types": [
+          {
+            "type": "unquoted_string",
+            "named": true
+          }
+        ]
+      },
+      "value": {
+        "multiple": false,
+        "required": true,
+        "types": [
+          {
+            "type": "double_quoted_string",
+            "named": true
+          },
+          {
+            "type": "unquoted_string",
+            "named": true
+          }
+        ]
+      }
+    }
+  },
+  {
+    "type": "expose_instruction",
+    "named": true,
+    "fields": {},
+    "children": {
+      "multiple": true,
+      "required": true,
+      "types": [
+        {
+          "type": "expose_port",
+          "named": true
+        }
+      ]
+    }
+  },
+  {
+    "type": "expose_port",
+    "named": true,
+    "fields": {}
+  },
+  {
+    "type": "from_instruction",
+    "named": true,
+    "fields": {},
+    "children": {
+      "multiple": true,
+      "required": true,
+      "types": [
+        {
+          "type": "image_spec",
+          "named": true
+        },
+        {
+          "type": "name",
+          "named": true
+        },
+        {
+          "type": "param",
+          "named": true
+        }
+      ]
+    }
+  },
+  {
+    "type": "healthcheck_instruction",
+    "named": true,
+    "fields": {},
+    "children": {
+      "multiple": true,
+      "required": false,
+      "types": [
+        {
+          "type": "cmd_instruction",
+          "named": true
+        },
+        {
+          "type": "param",
+          "named": true
+        }
+      ]
+    }
+  },
+  {
+    "type": "image_spec",
+    "named": true,
+    "fields": {},
+    "children": {
+      "multiple": true,
+      "required": false,
+      "types": [
+        {
+          "type": "image_digest",
+          "named": true
+        },
+        {
+          "type": "image_tag",
+          "named": true
+        }
+      ]
+    }
+  },
+  {
+    "type": "label_instruction",
+    "named": true,
+    "fields": {},
+    "children": {
+      "multiple": true,
+      "required": true,
+      "types": [
+        {
+          "type": "label_pair",
+          "named": true
+        }
+      ]
+    }
+  },
+  {
+    "type": "label_pair",
+    "named": true,
+    "fields": {
+      "key": {
+        "multiple": false,
+        "required": true,
+        "types": [
+          {
+            "type": "unquoted_string",
+            "named": true
+          }
+        ]
+      },
+      "value": {
+        "multiple": false,
+        "required": true,
+        "types": [
+          {
+            "type": "double_quoted_string",
+            "named": true
+          },
+          {
+            "type": "unquoted_string",
+            "named": true
+          }
+        ]
+      }
+    }
+  },
+  {
+    "type": "onbuild_instruction",
+    "named": true,
+    "fields": {},
+    "children": {
+      "multiple": false,
+      "required": true,
+      "types": [
+        {
+          "type": "add_instruction",
+          "named": true
+        },
+        {
+          "type": "arg_instruction",
+          "named": true
+        },
+        {
+          "type": "cmd_instruction",
+          "named": true
+        },
+        {
+          "type": "copy_instruction",
+          "named": true
+        },
+        {
+          "type": "entrypoint_instruction",
+          "named": true
+        },
+        {
+          "type": "env_instruction",
+          "named": true
+        },
+        {
+          "type": "expose_instruction",
+          "named": true
+        },
+        {
+          "type": "from_instruction",
+          "named": true
+        },
+        {
+          "type": "healthcheck_instruction",
+          "named": true
+        },
+        {
+          "type": "label_instruction",
+          "named": true
+        },
+        {
+          "type": "onbuild_instruction",
+          "named": true
+        },
+        {
+          "type": "run_instruction",
+          "named": true
+        },
+        {
+          "type": "shell_instruction",
+          "named": true
+        },
+        {
+          "type": "stopsignal_instruction",
+          "named": true
+        },
+        {
+          "type": "user_instruction",
+          "named": true
+        },
+        {
+          "type": "volume_instruction",
+          "named": true
+        },
+        {
+          "type": "workdir_instruction",
+          "named": true
+        }
+      ]
+    }
+  },
+  {
+    "type": "param",
+    "named": true,
+    "fields": {}
+  },
+  {
+    "type": "run_instruction",
+    "named": true,
+    "fields": {},
+    "children": {
+      "multiple": false,
+      "required": true,
+      "types": [
+        {
+          "type": "shell_command",
+          "named": true
+        },
+        {
+          "type": "string_array",
+          "named": true
+        }
+      ]
+    }
+  },
+  {
+    "type": "shell_command",
+    "named": true,
+    "fields": {}
+  },
+  {
+    "type": "shell_instruction",
+    "named": true,
+    "fields": {},
+    "children": {
+      "multiple": false,
+      "required": true,
+      "types": [
+        {
+          "type": "string_array",
+          "named": true
+        }
+      ]
+    }
+  },
+  {
+    "type": "source_file",
+    "named": true,
+    "fields": {},
+    "children": {
+      "multiple": true,
+      "required": false,
+      "types": [
+        {
+          "type": "add_instruction",
+          "named": true
+        },
+        {
+          "type": "arg_instruction",
+          "named": true
+        },
+        {
+          "type": "cmd_instruction",
+          "named": true
+        },
+        {
+          "type": "copy_instruction",
+          "named": true
+        },
+        {
+          "type": "entrypoint_instruction",
+          "named": true
+        },
+        {
+          "type": "env_instruction",
+          "named": true
+        },
+        {
+          "type": "expose_instruction",
+          "named": true
+        },
+        {
+          "type": "from_instruction",
+          "named": true
+        },
+        {
+          "type": "healthcheck_instruction",
+          "named": true
+        },
+        {
+          "type": "label_instruction",
+          "named": true
+        },
+        {
+          "type": "onbuild_instruction",
+          "named": true
+        },
+        {
+          "type": "run_instruction",
+          "named": true
+        },
+        {
+          "type": "shell_instruction",
+          "named": true
+        },
+        {
+          "type": "stopsignal_instruction",
+          "named": true
+        },
+        {
+          "type": "user_instruction",
+          "named": true
+        },
+        {
+          "type": "volume_instruction",
+          "named": true
+        },
+        {
+          "type": "workdir_instruction",
+          "named": true
+        }
+      ]
+    }
+  },
+  {
+    "type": "stopsignal_instruction",
+    "named": true,
+    "fields": {}
+  },
+  {
+    "type": "string_array",
+    "named": true,
+    "fields": {},
+    "children": {
+      "multiple": true,
+      "required": false,
+      "types": [
+        {
+          "type": "double_quoted_string",
+          "named": true
+        }
+      ]
+    }
+  },
+  {
+    "type": "unquoted_string",
+    "named": true,
+    "fields": {}
+  },
+  {
+    "type": "user_instruction",
+    "named": true,
+    "fields": {
+      "group": {
+        "multiple": false,
+        "required": false,
+        "types": [
+          {
+            "type": "unquoted_string",
+            "named": true
+          }
+        ]
+      },
+      "user": {
+        "multiple": false,
+        "required": true,
+        "types": [
+          {
+            "type": "unquoted_string",
+            "named": true
+          }
+        ]
+      }
+    }
+  },
+  {
+    "type": "volume_instruction",
+    "named": true,
+    "fields": {},
+    "children": {
+      "multiple": true,
+      "required": true,
+      "types": [
+        {
+          "type": "path",
+          "named": true
+        },
+        {
+          "type": "string_array",
+          "named": true
+        }
+      ]
+    }
+  },
+  {
+    "type": "workdir_instruction",
+    "named": true,
+    "fields": {},
+    "children": {
+      "multiple": false,
+      "required": true,
+      "types": [
+        {
+          "type": "path",
+          "named": true
+        }
+      ]
+    }
+  },
+  {
+    "type": "\n",
+    "named": false
+  },
+  {
+    "type": "\"",
+    "named": false
+  },
+  {
+    "type": "#",
+    "named": false
+  },
+  {
+    "type": ",",
+    "named": false
+  },
+  {
+    "type": "--",
+    "named": false
+  },
+  {
+    "type": "/tcp",
+    "named": false
+  },
+  {
+    "type": "/udp",
+    "named": false
+  },
+  {
+    "type": ":",
+    "named": false
+  },
+  {
+    "type": "=",
+    "named": false
+  },
+  {
+    "type": "ADD",
+    "named": false
+  },
+  {
+    "type": "ARG",
+    "named": false
+  },
+  {
+    "type": "AS",
+    "named": false
+  },
+  {
+    "type": "CMD",
+    "named": false
+  },
+  {
+    "type": "COPY",
+    "named": false
+  },
+  {
+    "type": "ENTRYPOINT",
+    "named": false
+  },
+  {
+    "type": "ENV",
+    "named": false
+  },
+  {
+    "type": "EXPOSE",
+    "named": false
+  },
+  {
+    "type": "FROM",
+    "named": false
+  },
+  {
+    "type": "HEALTHCHECK",
+    "named": false
+  },
+  {
+    "type": "LABEL",
+    "named": false
+  },
+  {
+    "type": "NONE",
+    "named": false
+  },
+  {
+    "type": "ONBUILD",
+    "named": false
+  },
+  {
+    "type": "RUN",
+    "named": false
+  },
+  {
+    "type": "SHELL",
+    "named": false
+  },
+  {
+    "type": "STOPSIGNAL",
+    "named": false
+  },
+  {
+    "type": "USER",
+    "named": false
+  },
+  {
+    "type": "VOLUME",
+    "named": false
+  },
+  {
+    "type": "WORKDIR",
+    "named": false
+  },
+  {
+    "type": "[",
+    "named": false
+  },
+  {
+    "type": "\\\n",
+    "named": false
+  },
+  {
+    "type": "\\ ",
+    "named": false
+  },
+  {
+    "type": "]",
+    "named": false
+  },
+  {
+    "type": "escape_sequence",
+    "named": true
+  },
+  {
+    "type": "image_digest",
+    "named": true
+  },
+  {
+    "type": "image_tag",
+    "named": true
+  },
+  {
+    "type": "name",
+    "named": false
+  },
+  {
+    "type": "name",
+    "named": true
+  },
+  {
+    "type": "path",
+    "named": true
+  }
+]

src/tree_sitter/parser.h

@@ -0,0 +1,223 @@
+#ifndef TREE_SITTER_PARSER_H_
+#define TREE_SITTER_PARSER_H_
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#include <stdbool.h>
+#include <stdint.h>
+#include <stdlib.h>
+
+#define ts_builtin_sym_error ((TSSymbol)-1)
+#define ts_builtin_sym_end 0
+#define TREE_SITTER_SERIALIZATION_BUFFER_SIZE 1024
+
+typedef uint16_t TSStateId;
+
+#ifndef TREE_SITTER_API_H_
+typedef uint16_t TSSymbol;
+typedef uint16_t TSFieldId;
+typedef struct TSLanguage TSLanguage;
+#endif
+
+typedef struct {
+  TSFieldId field_id;
+  uint8_t child_index;
+  bool inherited;
+} TSFieldMapEntry;
+
+typedef struct {
+  uint16_t index;
+  uint16_t length;
+} TSFieldMapSlice;
+
+typedef struct {
+  bool visible;
+  bool named;
+  bool supertype;
+} TSSymbolMetadata;
+
+typedef struct TSLexer TSLexer;
+
+struct TSLexer {
+  int32_t lookahead;
+  TSSymbol result_symbol;
+  void (*advance)(TSLexer *, bool);
+  void (*mark_end)(TSLexer *);
+  uint32_t (*get_column)(TSLexer *);
+  bool (*is_at_included_range_start)(const TSLexer *);
+  bool (*eof)(const TSLexer *);
+};
+
+typedef enum {
+  TSParseActionTypeShift,
+  TSParseActionTypeReduce,
+  TSParseActionTypeAccept,
+  TSParseActionTypeRecover,
+} TSParseActionType;
+
+typedef union {
+  struct {
+    uint8_t type;
+    TSStateId state;
+    bool extra;
+    bool repetition;
+  } shift;
+  struct {
+    uint8_t type;
+    uint8_t child_count;
+    TSSymbol symbol;
+    int16_t dynamic_precedence;
+    uint16_t production_id;
+  } reduce;
+  uint8_t type;
+} TSParseAction;
+
+typedef struct {
+  uint16_t lex_state;
+  uint16_t external_lex_state;
+} TSLexMode;
+
+typedef union {
+  TSParseAction action;
+  struct {
+    uint8_t count;
+    bool reusable;
+  } entry;
+} TSParseActionEntry;
+
+struct TSLanguage {
+  uint32_t version;
+  uint32_t symbol_count;
+  uint32_t alias_count;
+  uint32_t token_count;
+  uint32_t external_token_count;
+  uint32_t state_count;
+  uint32_t large_state_count;
+  uint32_t production_id_count;
+  uint32_t field_count;
+  uint16_t max_alias_sequence_length;
+  const uint16_t *parse_table;
+  const uint16_t *small_parse_table;
+  const uint32_t *small_parse_table_map;
+  const TSParseActionEntry *parse_actions;
+  const char **symbol_names;
+  const char **field_names;
+  const TSFieldMapSlice *field_map_slices;
+  const TSFieldMapEntry *field_map_entries;
+  const TSSymbolMetadata *symbol_metadata;
+  const TSSymbol *public_symbol_map;
+  const uint16_t *alias_map;
+  const TSSymbol *alias_sequences;
+  const TSLexMode *lex_modes;
+  bool (*lex_fn)(TSLexer *, TSStateId);
+  bool (*keyword_lex_fn)(TSLexer *, TSStateId);
+  TSSymbol keyword_capture_token;
+  struct {
+    const bool *states;
+    const TSSymbol *symbol_map;
+    void *(*create)(void);
+    void (*destroy)(void *);
+    bool (*scan)(void *, TSLexer *, const bool *symbol_whitelist);
+    unsigned (*serialize)(void *, char *);
+    void (*deserialize)(void *, const char *, unsigned);
+  } external_scanner;
+};
+
+/*
+ *  Lexer Macros
+ */
+
+#define START_LEXER()           \
+  bool result = false;          \
+  bool skip = false;            \
+  bool eof = false;             \
+  int32_t lookahead;            \
+  goto start;                   \
+  next_state:                   \
+  lexer->advance(lexer, skip);  \
+  start:                        \
+  skip = false;                 \
+  lookahead = lexer->lookahead;
+
+#define ADVANCE(state_value) \
+  {                          \
+    state = state_value;     \
+    goto next_state;         \
+  }
+
+#define SKIP(state_value) \
+  {                       \
+    skip = true;          \
+    state = state_value;  \
+    goto next_state;      \
+  }
+
+#define ACCEPT_TOKEN(symbol_value)     \
+  result = true;                       \
+  lexer->result_symbol = symbol_value; \
+  lexer->mark_end(lexer);
+
+#define END_STATE() return result;
+
+/*
+ *  Parse Table Macros
+ */
+
+#define SMALL_STATE(id) id - LARGE_STATE_COUNT
+
+#define STATE(id) id
+
+#define ACTIONS(id) id
+
+#define SHIFT(state_value)            \
+  {{                                  \
+    .shift = {                        \
+      .type = TSParseActionTypeShift, \
+      .state = state_value            \
+    }                                 \
+  }}
+
+#define SHIFT_REPEAT(state_value)     \
+  {{                                  \
+    .shift = {                        \
+      .type = TSParseActionTypeShift, \
+      .state = state_value,           \
+      .repetition = true              \
+    }                                 \
+  }}
+
+#define SHIFT_EXTRA()                 \
+  {{                                  \
+    .shift = {                        \
+      .type = TSParseActionTypeShift, \
+      .extra = true                   \
+    }                                 \
+  }}
+
+#define REDUCE(symbol_val, child_count_val, ...) \
+  {{                                             \
+    .reduce = {                                  \
+      .type = TSParseActionTypeReduce,           \
+      .symbol = symbol_val,                      \
+      .child_count = child_count_val,            \
+      __VA_ARGS__                                \
+    },                                           \
+  }}
+
+#define RECOVER()                    \
+  {{                                 \
+    .type = TSParseActionTypeRecover \
+  }}
+
+#define ACCEPT_INPUT()              \
+  {{                                \
+    .type = TSParseActionTypeAccept \
+  }}
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif  // TREE_SITTER_PARSER_H_