implement smarty comments by extending html

Author: deltragon

.gitignore

@@ -0,0 +1,5 @@
+node_modules
+build
+Cargo.lock
+package-lock.json
+/target/

Cargo.toml

@@ -0,0 +1,26 @@
+[package]
+name = "tree-sitter-htmlsmarty"
+description = "htmlsmarty grammar for the tree-sitter parsing library"
+version = "0.0.1"
+keywords = ["incremental", "parsing", "htmlsmarty"]
+categories = ["parsing", "text-editors"]
+repository = "https://github.com/tree-sitter/tree-sitter-htmlsmarty"
+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 = "~0.20.10"
+
+[build-dependencies]
+cc = "1.0"

binding.gyp

@@ -0,0 +1,19 @@
+{
+  "targets": [
+    {
+      "target_name": "tree_sitter_htmlsmarty_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_htmlsmarty();
+
+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_htmlsmarty());
+
+  Nan::Set(instance, Nan::New("name").ToLocalChecked(), Nan::New("htmlsmarty").ToLocalChecked());
+  Nan::Set(module, Nan::New("exports").ToLocalChecked(), instance);
+}
+
+NODE_MODULE(tree_sitter_htmlsmarty_binding, Init)
+
+}  // namespace

bindings/node/index.js

@@ -0,0 +1,19 @@
+try {
+  module.exports = require("../../build/Release/tree_sitter_htmlsmarty_binding");
+} catch (error1) {
+  if (error1.code !== 'MODULE_NOT_FOUND') {
+    throw error1;
+  }
+  try {
+    module.exports = require("../../build/Debug/tree_sitter_htmlsmarty_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 +1,38 @@
+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 htmlsmarty 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_htmlsmarty::language()).expect("Error loading htmlsmarty 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_htmlsmarty() -> 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_htmlsmarty() }
+}
+
+/// 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 htmlsmarty language");
+    }
+}

grammar.js

@@ -0,0 +1,87 @@
+const html = require("tree-sitter-html/grammar");
+
+module.exports = grammar(html, {
+  name: "htmlsmarty",
+
+  extras: $ => [
+    $.comment,
+    // TODO: technically this would be possible too, and a whole lot easier (only needs adjusting text, nothing else)
+    // but i dont know if this would then properly work with indentation for eg. smarty_if
+    //$.smarty_comment,
+    /\s+/,
+  ],
+
+  rules: {
+    smarty_comment: ($) => seq(
+      "{*",
+      /[^*]*\*+([^}*][^*]*\*+)*/,
+      "}"
+    ),
+
+    // in text
+    // FIXME: see comment in test/corpus/unpaired.txt for why this is broken in the edge case of
+    // `text {<valid-tag/>`
+    text: $ => /([^<>{]|\{[^*<>])+/,
+
+    _node: $ => choice(
+      $.doctype,
+      $.text,
+      $.smarty_comment,
+      $.element,
+      $.script_element,
+      $.style_element,
+      $.erroneous_end_tag
+    ),
+
+    // in attribute lists
+    _attribute: $ => choice(
+      $.attribute,
+      $.smarty_comment,
+    ),
+
+    attribute_name: $ => /([^<>"'/=\s{]|\{[^*<>"'/=\s{])+/,
+
+    sq_attribute_value_fragment: $ => repeat1(choice(
+      /([^'{]|\{[^*'])+/,
+      $.smarty_comment
+    )),
+    dq_attribute_value_fragment: $ => repeat1(choice(
+      /([^"{]|\{[^*"])+/,
+      $.smarty_comment
+    )),
+
+    quoted_attribute_value: $ => choice(
+      seq("'", optional(alias($.sq_attribute_value_fragment, $.attribute_value)), optional("{"), "'"),
+      seq('"', optional(alias($.dq_attribute_value_fragment, $.attribute_value)), optional("{"), '"'),
+    ),
+
+
+    start_tag: $ => seq(
+      '<',
+      alias($._start_tag_name, $.tag_name),
+      repeat($._attribute),
+      '>'
+    ),
+
+    script_start_tag: $ => seq(
+      '<',
+      alias($._script_start_tag_name, $.tag_name),
+      repeat($._attribute),
+      '>'
+    ),
+
+    style_start_tag: $ => seq(
+      '<',
+      alias($._style_start_tag_name, $.tag_name),
+      repeat($._attribute),
+      '>'
+    ),
+
+    self_closing_tag: $ => seq(
+      '<',
+      alias($._start_tag_name, $.tag_name),
+      repeat($._attribute),
+      '/>'
+    ),
+  }
+});

package.json

@@ -0,0 +1,29 @@
+{
+  "name": "tree-sitter-htmlsmarty",
+  "version": "0.1.0",
+  "description": "A tree-sitter grammar for the smarty template language",
+  "main": "bindings/node",
+  "scripts": {
+    "test": "tree-sitter test"
+  },
+  "keywords": [
+    "smarty",
+    "html",
+    "tree-sitter",
+    "parser",
+    "grammar"
+  ],
+  "author": "deltragon",
+  "license": "MIT",
+  "dependencies": {
+    "nan": "^2.15.0",
+    "tree-sitter-html": "0.19.0"
+  },
+  "repository": {
+    "type": "git",
+    "url": "git://github.com/deltragon/tree-sitter-htmlsmarty.git"
+  },
+  "devDependencies": {
+    "tree-sitter-cli": "^0.20.4"
+  }
+}