Add support for TOML

CHANGELOG.md

@@ -2,7 +2,7 @@
 
 ### Parsing
 
-Added support for Kotlin.
+Added support for Kotlin and TOML.
 
 Fixed an issue with YAML and `|` block strings.
 

build.rs

@@ -194,6 +194,11 @@ fn main() {
             src_dir: "vendor/tree-sitter-scala-src",
             extra_files: vec!["scanner.c"],
         },
+        TreeSitterParser {
+            name: "tree-sitter-toml",
+            src_dir: "vendor/tree-sitter-toml-src",
+            extra_files: vec!["scanner.c"],
+        },
         TreeSitterParser {
             name: "tree-sitter-tsx",
             src_dir: "vendor/tree-sitter-typescript-src/tsx/src",

manual/src/upstream_parsers.md

@@ -30,6 +30,7 @@ Difftastic uses the following tree-sitter parsers:
 | Ruby            | [tree-sitter/tree-sitter-ruby](https://github.com/tree-sitter/tree-sitter-ruby)                                                                       |
 | Rust            | [tree-sitter/tree-sitter-rust](https://github.com/tree-sitter/tree-sitter-rust) ([forked](https://github.com/Wilfred/tree-sitter-rust))               |
 | Scala           | [tree-sitter/tree-sitter-scala](https://github.com/tree-sitter/tree-sitter-scala)                                                                     |
+| TOML            | [ikatyang/tree-sitter-toml](https://github.com/ikatyang/tree-sitter-toml)                                                                             |
 | TypeScript, TSX | [tree-sitter/tree-sitter-typescript](https://github.com/tree-sitter/tree-sitter-typescript)                                                           |
 | YAML            | [ikatyang/tree-sitter-yaml](https://github.com/ikatyang/tree-sitter-yaml)                                                                             |
 | Zig             | [maxxnino/tree-sitter-zig](https://github.com/maxxnino/tree-sitter-zig)                                                                               |

sample_files/compare.expected

@@ -139,6 +139,9 @@ dfc3495b8d5931029b479f0c878a3219  -
 sample_files/todomvc_before.gleam sample_files/todomvc_after.gleam
 c1d8b44875121d81c583dd3a8fb43232  -
 
+sample_files/toml_before.toml sample_files/toml_after.toml
+9928c1f07b655860c27e2e1002d94106  -
+
 sample_files/typing_before.ml sample_files/typing_after.ml
 ceba89e4ffc8406454d337638c7d45e6  -
 

sample_files/toml_after.toml

@@ -0,0 +1,27 @@
+# This is a TOML document
+
+title = "TOML Example Changed"
+
+[owner]
+name = "Tom Preston-Werner"
+dob = 2000-01-31T07:32:00-08:00
+
+[database]
+enabled = true
+ports = [ 8000, 8002 ]
+data = [ ["delta", "phi"], [3.14] ]
+temp_targets = { cpu = 79.5, case = 72.0 }
+
+[servers]
+
+[servers.alpha]
+ip = "10.0.0.1"
+role = "frontend"
+
+[servers.beta]
+ip = "10.0.0.2"
+role = "backend"
+
+str1 = """
+Roses are red
+Violets are blue"""

sample_files/toml_before.toml

@@ -0,0 +1,35 @@
+# This is a TOML document
+
+title = "TOML Example"
+
+[owner]
+name = "Tom Preston-Werner"
+dob = 1979-05-27T07:32:00-08:00
+
+[database]
+enabled = true
+ports = [ 8000, 8001, 8002 ]
+data = [ ["delta", "phi"], [3.14] ]
+temp_targets = { cpu = 79.5, case = 72.0 }
+
+[servers]
+
+[servers.alpha]
+ip = "10.0.0.1"
+role = "frontend"
+
+[servers.beta]
+ip = "10.0.0.2"
+role = "backend"
+
+str1 = """
+Roses are red
+Violets are blue"""
+
+str2 = """\
+  The quick brown \
+  fox jumps over \
+  the lazy dog.\
+  """
+
+path = 'C:\Users\nodejs\templates'

src/guess_language.rs

@@ -47,6 +47,7 @@ pub enum Language {
     Ruby,
     Rust,
     Scala,
+    Toml,
     Tsx,
     TypeScript,
     Yaml,
@@ -117,6 +118,7 @@ fn from_emacs_mode_header(src: &str) -> Option<Language> {
             "rust" => Some(Rust),
             "scala" => Some(Scala),
             "sh" => Some(Bash),
+            "toml" => Some(Toml),
             "tuareg" => Some(OCaml),
             "typescript" => Some(TypeScript),
             "yaml" => Some(Yaml),
@@ -175,6 +177,7 @@ fn from_name(path: &Path) -> Option<Language> {
             ".emacs" | "_emacs" | "Cask" => Some(EmacsLisp),
             "TARGETS" | "BUCK" | "DEPS" => Some(Python),
             "Gemfile" | "Rakefile" => Some(Ruby),
+            "Cargo.lock" | "Gopkg.lock" | "Pipfile" | "poetry.lock" => Some(Toml),
             _ => None,
         },
         None => None,
@@ -221,6 +224,7 @@ pub fn from_extension(extension: &OsStr) -> Option<Language> {
         "rb" | "builder" | "spec" | "rake" => Some(Ruby),
         "rs" => Some(Rust),
         "scala" | "sbt" | "sc" => Some(Scala),
+        "toml" => Some(Toml),
         "ts" => Some(TypeScript),
         "tsx" => Some(Tsx),
         "yaml" | "yml" => Some(Yaml),

src/sliders.rs

@@ -54,9 +54,9 @@ fn prefer_outer_delimiter(language: guess_language::Language) -> bool {
         // For Lisp family languages, we get the best result with the
         // outer delimiter.
         EmacsLisp | Clojure | CommonLisp | JanetSimple => true,
-        // JSON is like Lisp: the outer delimiter in an array object
+        // JSON and TOML are like Lisp: the outer delimiter in an array object
         // is the most relevant.
-        Json => true,
+        Json | Toml => true,
         // For everything else, prefer the inner delimiter. These
         // languages have syntax like `foo(bar)` or `foo[bar]` where
         // the inner delimiter is more relevant.

src/tree_sitter_parser.rs

@@ -71,6 +71,7 @@ extern "C" {
     fn tree_sitter_ruby() -> ts::Language;
     fn tree_sitter_rust() -> ts::Language;
     fn tree_sitter_scala() -> ts::Language;
+    fn tree_sitter_toml() -> ts::Language;
     fn tree_sitter_tsx() -> ts::Language;
     fn tree_sitter_typescript() -> ts::Language;
     fn tree_sitter_yaml() -> ts::Language;
@@ -538,6 +539,20 @@ pub fn from_language(language: guess::Language) -> TreeSitterConfig {
                 .unwrap(),
             }
         }
+        Toml => {
+            let language = unsafe { tree_sitter_toml() };
+            TreeSitterConfig {
+                name: "TOML",
+                language,
+                atom_nodes: vec!["string"].into_iter().collect(),
+                delimiter_tokens: vec![("{", "}"), ("[", "]")],
+                highlight_query: ts::Query::new(
+                    language,
+                    include_str!("../vendor/highlights/toml.scm"),
+                )
+                .unwrap(),
+            }
+        }
         TypeScript => {
             let language = unsafe { tree_sitter_typescript() };
             TreeSitterConfig {

vendor/highlights/toml.scm

@@ -0,0 +1 @@
+../tree-sitter-toml/queries/highlights.scm

vendor/tree-sitter-toml-src

@@ -0,0 +1 @@
+tree-sitter-toml/src