Detect filetype from shebang line (#1001)

book/src/guides/adding_languages.md

@@ -33,10 +33,11 @@ ## languages.toml
 | scope         | A string like `source.js` that identifies the language. Currently, we strive to match the scope names used by popular TextMate grammars and by the Linguist library. Usually `source.<name>` or `text.<name>` in case of markup languages |
 | injection-regex | regex pattern that will be tested against a language name in order to determine whether this language should be used for a potential [language injection][treesitter-language-injection] site. |
 | file-types    | The filetypes of the language, for example `["yml", "yaml"]`  |
+| shebangs      | The interpreters from the shebang line, for example `["sh", "bash"]` |
 | roots         | A set of marker files to look for when trying to find the workspace root. For example `Cargo.lock`, `yarn.lock` |
 | auto-format   | Whether to autoformat this language when saving               |
 | comment-token | The token to use as a comment-token                           |
-| indent        | The indent to use. Has sub keys `tab-width` and `unit`            |
+| indent        | The indent to use. Has sub keys `tab-width` and `unit`        |
 | config        | Language server configuration                                 |
 
 ## Queries

helix-core/src/indent.rs

@@ -450,6 +450,7 @@ pub fn change<I>(document: &Document, changes: I) -> Self
             language: vec![LanguageConfiguration {
                 scope: "source.rust".to_string(),
                 file_types: vec!["rs".to_string()],
+                shebangs: vec![],
                 language_id: "Rust".to_string(),
                 highlight_config: OnceCell::new(),
                 config: None,

helix-core/src/syntax.rs

@@ -14,6 +14,8 @@
     cell::RefCell,
     collections::{HashMap, HashSet},
     fmt,
+    fs::File,
+    io::Read,
     path::Path,
     sync::Arc,
 };
@@ -52,6 +54,7 @@ pub struct LanguageConfiguration {
     pub language_id: String,
     pub scope: String,           // source.rust
     pub file_types: Vec<String>, // filename ends_with? <Gemfile, rb, etc>
+    pub shebangs: Vec<String>,   // interpreter(s) associated with language
     pub roots: Vec<String>,      // these indicate project roots <.git, Cargo.toml>
     pub comment_token: Option<String>,
 
@@ -254,6 +257,7 @@ pub struct Loader {
     // highlight_names ?
     language_configs: Vec<Arc<LanguageConfiguration>>,
     language_config_ids_by_file_type: HashMap<String, usize>, // Vec<usize>
+    language_config_ids_by_shebang: HashMap<String, usize>,
 }
 
 impl Loader {
@@ -261,6 +265,7 @@ pub fn new(config: Configuration) -> Self {
         let mut loader = Self {
             language_configs: Vec::new(),
             language_config_ids_by_file_type: HashMap::new(),
+            language_config_ids_by_shebang: HashMap::new(),
         };
 
         for config in config.language {
@@ -273,6 +278,11 @@ pub fn new(config: Configuration) -> Self {
                     .language_config_ids_by_file_type
                     .insert(file_type.clone(), language_id);
             }
+            for shebang in &config.shebangs {
+                loader
+                    .language_config_ids_by_shebang
+                    .insert(shebang.clone(), language_id);
+            }
 
             loader.language_configs.push(Arc::new(config));
         }
@@ -298,6 +308,20 @@ pub fn language_config_for_file_name(&self, path: &Path) -> Option<Arc<LanguageC
         // TODO: content_regex handling conflict resolution
     }
 
+    pub fn language_config_for_shebang(&self, path: &Path) -> Option<Arc<LanguageConfiguration>> {
+        // Read the first 128 bytes of the file. If its a shebang line, try to find the language
+        let file = File::open(path).ok()?;
+        let mut buf = String::with_capacity(128);
+        file.take(128).read_to_string(&mut buf).ok()?;
+        static SHEBANG_REGEX: Lazy<Regex> =
+            Lazy::new(|| Regex::new(r"^#!\s*(?:\S*[/\\](?:env\s+)?)?([^\s\.\d]+)").unwrap());
+        let configuration_id = SHEBANG_REGEX
+            .captures(&buf)
+            .and_then(|cap| self.language_config_ids_by_shebang.get(&cap[1]));
+
+        configuration_id.and_then(|&id| self.language_configs.get(id).cloned())
+    }
+
     pub fn language_config_for_scope(&self, scope: &str) -> Option<Arc<LanguageConfiguration>> {
         self.language_configs
             .iter()

helix-view/src/document.rs

@@ -494,7 +494,9 @@ fn save_impl<F: Future<Output = LspFormatting>>(
     /// Detect the programming language based on the file type.
     pub fn detect_language(&mut self, theme: Option<&Theme>, config_loader: &syntax::Loader) {
         if let Some(path) = &self.path {
-            let language_config = config_loader.language_config_for_file_name(path);
+            let language_config = config_loader
+                .language_config_for_file_name(path)
+                .or_else(|| config_loader.language_config_for_shebang(path));
             self.set_language(theme, language_config);
         }
     }

languages.toml

@@ -3,6 +3,7 @@ name = "rust"
 scope = "source.rust"
 injection-regex = "rust"
 file-types = ["rs"]
+shebangs = []
 roots = []
 auto-format = true
 comment-token = "//"
@@ -17,6 +18,7 @@ name = "toml"
 scope = "source.toml"
 injection-regex = "toml"
 file-types = ["toml"]
+shebangs = []
 roots = []
 comment-token = "#"
 
@@ -27,6 +29,7 @@ name = "protobuf"
 scope = "source.proto"
 injection-regex = "protobuf"
 file-types = ["proto"]
+shebangs = []
 roots = []
 comment-token = "//"
 
@@ -37,6 +40,7 @@ name = "elixir"
 scope = "source.elixir"
 injection-regex = "elixir"
 file-types = ["ex", "exs"]
+shebangs = []
 roots = []
 comment-token = "#"
 
@@ -48,6 +52,7 @@ name = "mint"
 scope = "source.mint"
 injection-regex = "mint"
 file-types = ["mint"]
+shebangs = []
 roots = []
 comment-token = "//"
 
@@ -59,6 +64,7 @@ name = "json"
 scope = "source.json"
 injection-regex = "json"
 file-types = ["json"]
+shebangs = []
 roots = []
 
 indent = { tab-width = 2, unit = "  " }
@@ -68,6 +74,7 @@ name = "c"
 scope = "source.c"
 injection-regex = "c"
 file-types = ["c"] # TODO: ["h"]
+shebangs = []
 roots = []
 comment-token = "//"
 
@@ -79,6 +86,7 @@ name = "cpp"
 scope = "source.cpp"
 injection-regex = "cpp"
 file-types = ["cc", "hh", "cpp", "hpp", "h", "ipp", "tpp", "cxx", "hxx", "ixx", "txx", "ino"]
+shebangs = []
 roots = []
 comment-token = "//"
 
@@ -90,6 +98,7 @@ name = "c-sharp"
 scope = "source.csharp"
 injection-regex = "c-?sharp"
 file-types = ["cs"]
+shebangs = []
 roots = []
 comment-token = "//"
 
@@ -100,6 +109,7 @@ name = "go"
 scope = "source.go"
 injection-regex = "go"
 file-types = ["go"]
+shebangs = []
 roots = ["Gopkg.toml", "go.mod"]
 auto-format = true
 comment-token = "//"
@@ -113,6 +123,7 @@ name = "javascript"
 scope = "source.js"
 injection-regex = "^(js|javascript)$"
 file-types = ["js", "mjs"]
+shebangs = []
 roots = []
 comment-token = "//"
 # TODO: highlights-jsx, highlights-params
@@ -124,6 +135,7 @@ name = "typescript"
 scope = "source.ts"
 injection-regex = "^(ts|typescript)$"
 file-types = ["ts"]
+shebangs = []
 roots = []
 # TODO: highlights-jsx, highlights-params
 
@@ -135,6 +147,7 @@ name = "tsx"
 scope = "source.tsx"
 injection-regex = "^(tsx)$" # |typescript
 file-types = ["tsx"]
+shebangs = []
 roots = []
 # TODO: highlights-jsx, highlights-params
 
@@ -146,6 +159,7 @@ name = "css"
 scope = "source.css"
 injection-regex = "css"
 file-types = ["css"]
+shebangs = []
 roots = []
 
 indent = { tab-width = 2, unit = "  " }
@@ -155,6 +169,7 @@ name = "html"
 scope = "text.html.basic"
 injection-regex = "html"
 file-types = ["html"]
+shebangs = []
 roots = []
 
 indent = { tab-width = 2, unit = "  " }
@@ -164,6 +179,7 @@ name = "python"
 scope = "source.python"
 injection-regex = "python"
 file-types = ["py"]
+shebangs = ["python"]
 roots = []
 comment-token = "#"
 
@@ -176,6 +192,7 @@ name = "nix"
 scope = "source.nix"
 injection-regex = "nix"
 file-types = ["nix"]
+shebangs = []
 roots = []
 comment-token = "#"
 
@@ -187,6 +204,7 @@ name = "ruby"
 scope = "source.ruby"
 injection-regex = "ruby"
 file-types = ["rb"]
+shebangs = ["ruby"]
 roots = []
 comment-token = "#"
 
@@ -198,6 +216,7 @@ name = "bash"
 scope = "source.bash"
 injection-regex = "bash"
 file-types = ["sh", "bash"]
+shebangs = ["sh", "bash", "dash"]
 roots = []
 comment-token = "#"
 
@@ -209,6 +228,7 @@ name = "php"
 scope = "source.php"
 injection-regex = "php"
 file-types = ["php"]
+shebangs = ["php"]
 roots = []
 
 indent = { tab-width = 4, unit = "    " }
@@ -218,6 +238,7 @@ name = "latex"
 scope = "source.tex"
 injection-regex = "tex"
 file-types = ["tex"]
+shebangs = []
 roots = []
 comment-token = "%"
 
@@ -228,6 +249,7 @@ name = "julia"
 scope = "source.julia"
 injection-regex = "julia"
 file-types = ["jl"]
+shebangs = []
 roots = []
 comment-token = "#"
 language-server = { command = "julia", args = [
@@ -253,6 +275,7 @@ name = "java"
 scope = "source.java"
 injection-regex = "java"
 file-types = ["java"]
+shebangs = []
 roots = []
 indent = { tab-width = 4, unit = "    " }
 
@@ -261,6 +284,7 @@ name = "ledger"
 scope = "source.ledger"
 injection-regex = "ledger"
 file-types = ["ldg", "ledger", "journal"]
+shebangs = []
 roots = []
 comment-token = ";"
 indent = { tab-width = 4, unit = "    " }
@@ -270,6 +294,7 @@ name = "ocaml"
 scope = "source.ocaml"
 injection-regex = "ocaml"
 file-types = ["ml"]
+shebangs = []
 roots = []
 comment-token = "(**)"
 indent = { tab-width = 2, unit = "  " }
@@ -278,6 +303,7 @@ indent = { tab-width = 2, unit = "  " }
 name = "ocaml-interface"
 scope = "source.ocaml.interface"
 file-types = ["mli"]
+shebangs = []
 roots = []
 comment-token = "(**)"
 indent = { tab-width = 2, unit = "  "}
@@ -286,6 +312,7 @@ indent = { tab-width = 2, unit = "  "}
 name = "lua"
 scope = "source.lua"
 file-types = ["lua"]
+shebangs = []
 roots = []
 comment-token = "--"
 indent = { tab-width = 2, unit = "  " }
@@ -295,6 +322,7 @@ name = "svelte"
 scope = "source.svelte"
 injection-regex = "svelte"
 file-types = ["svelte"]
+shebangs = []
 roots = []
 indent = { tab-width = 2, unit = "  " }
 language-server = { command = "svelteserver", args = ["--stdio"] }
@@ -305,6 +333,7 @@ name = "vue"
 scope = "source.vue"
 injection-regex = "vue"
 file-types = ["vue"]
+shebangs = []
 roots = []
 indent = { tab-width = 2, unit = "  " }
 
@@ -312,6 +341,7 @@ indent = { tab-width = 2, unit = "  " }
 name = "yaml"
 scope = "source.yaml"
 file-types = ["yml", "yaml"]
+shebangs = []
 roots = []
 comment-token = "#"
 indent = { tab-width = 2, unit = "  " }
@@ -331,6 +361,7 @@ name = "zig"
 scope = "source.zig"
 injection-regex = "zig"
 file-types = ["zig"]
+shebangs = []
 roots = ["build.zig"]
 auto-format = true
 comment-token = "//"
@@ -343,6 +374,7 @@ name = "prolog"
 scope = "source.prolog"
 roots = []
 file-types = ["pl", "prolog"]
+shebangs = ["swipl"]
 comment-token = "%"
 
 language-server = { command = "swipl", args = [
@@ -354,6 +386,7 @@ language-server = { command = "swipl", args = [
 name = "tsq"
 scope = "source.tsq"
 file-types = ["scm"]
+shebangs = []
 roots = []
 comment-token = ";"
 indent = { tab-width = 2, unit = "  " }
@@ -362,6 +395,7 @@ indent = { tab-width = 2, unit = "  " }
 name = "cmake"
 scope = "source.cmake"
 file-types = ["cmake", "CMakeLists.txt"]
+shebangs = []
 roots = []
 comment-token = "#"
 indent = { tab-width = 2, unit = "  " }
@@ -371,6 +405,7 @@ language-server = { command = "cmake-language-server" }
 name = "perl"
 scope = "source.perl"
 file-types = ["pl", "pm"]
+shebangs = ["perl"]
 roots = []
 comment-token = "#"
 indent = { tab-width = 2, unit = "  " }