Refactor: Add built-in support for Rust, JS, Python

.editorconfig

@@ -0,0 +1,9 @@
+root = true
+
+[*]
+indent_style = space
+indent_size = 4
+end_of_line = lf
+charset = utf-8
+trim_trailing_whitespace = false
+insert_final_newline = true

Cargo.toml

@@ -3,7 +3,9 @@ name = "chroma-syntaxis"
 version = "0.1.0"
 edition = "2021"
 
-# See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
+[features]
+default = ["built-in"]
+built-in = ["tree-sitter-rust", "tree-sitter-javascript", "tree-sitter-regex", "tree-sitter-python"]
 
 [dependencies]
 html-escape = "0.2.11"
@@ -11,4 +13,8 @@ once_cell = "1.10.0"
 thiserror = "1.0.30"
 tree-sitter = "0.20.6"
 tree-sitter-highlight = "0.20.1"
-tree-sitter-rust = "0.20.1"
+
+tree-sitter-rust = { version = "0.20.1", optional = true }
+tree-sitter-javascript = { version = "0.20.0", optional = true }
+tree-sitter-regex = { git = "https://github.com/tree-sitter/tree-sitter-regex.git", rev = "e1cfca", optional = true }
+tree-sitter-python = { version = "0.19.1", optional = true }

samples/fizzbuzz.js

@@ -0,0 +1,17 @@
+const fizzbuzz = n => {
+    if (n % 15 == 0)
+        return "Fizz buzz";
+    if (n % 3 == 0)
+        return "Fizz";
+    if (n % 5 == 0)
+        return "Buzz";
+    return n;
+}
+
+const main = () => {
+    for (let n = 1; n <= 100; n++) {
+        console.log(fizzbuzz(n));
+    }
+};
+
+main();

samples/fizzbuzz.js.html

@@ -0,0 +1,17 @@
+<span class="keyword">const</span> <span class="function">fizzbuzz</span> <span class="operator">=</span> <span class="variable">n</span> <span class="operator">=&gt;</span> <span class="punctuation bracket">{</span>
+    <span class="keyword">if</span> <span class="punctuation bracket">(</span><span class="variable">n</span> <span class="operator">%</span> 15 <span class="operator">==</span> 0<span class="punctuation bracket">)</span>
+        <span class="keyword">return</span> <span class="string">"Fizz buzz"</span><span class="punctuation delimiter">;</span>
+    <span class="keyword">if</span> <span class="punctuation bracket">(</span><span class="variable">n</span> <span class="operator">%</span> 3 <span class="operator">==</span> 0<span class="punctuation bracket">)</span>
+        <span class="keyword">return</span> <span class="string">"Fizz"</span><span class="punctuation delimiter">;</span>
+    <span class="keyword">if</span> <span class="punctuation bracket">(</span><span class="variable">n</span> <span class="operator">%</span> 5 <span class="operator">==</span> 0<span class="punctuation bracket">)</span>
+        <span class="keyword">return</span> <span class="string">"Buzz"</span><span class="punctuation delimiter">;</span>
+    <span class="keyword">return</span> <span class="variable">n</span><span class="punctuation delimiter">;</span>
+<span class="punctuation bracket">}</span>
+
+<span class="keyword">const</span> <span class="function">main</span> <span class="operator">=</span> <span class="punctuation bracket">()</span> <span class="operator">=&gt;</span> <span class="punctuation bracket">{</span>
+    <span class="keyword">for</span> <span class="punctuation bracket">(</span><span class="keyword">let</span> <span class="variable">n</span> <span class="operator">=</span> 1<span class="punctuation delimiter">;</span> <span class="variable">n</span> <span class="operator">&lt;=</span> 100<span class="punctuation delimiter">;</span> <span class="variable">n</span><span class="operator">++</span><span class="punctuation bracket">)</span> <span class="punctuation bracket">{</span>
+        <span class="variable builtin">console</span><span class="punctuation delimiter">.</span><span class="function">log</span><span class="punctuation bracket">(</span><span class="function">fizzbuzz</span><span class="punctuation bracket">(</span><span class="variable">n</span><span class="punctuation bracket">))</span><span class="punctuation delimiter">;</span>
+    <span class="punctuation bracket">}</span>
+<span class="punctuation bracket">}</span><span class="punctuation delimiter">;</span>
+
+<span class="function">main</span><span class="punctuation bracket">()</span><span class="punctuation delimiter">;</span>

samples/fizzbuzz.py

@@ -0,0 +1,8 @@
+def main():
+    for n in range(1, 101):
+        lut = [n, "Fizz", "Buzz", "Fizz Buzz"]
+        idx = (n % 3 == 0) + 2 * (n % 5 == 0)
+        print(lut[idx])
+
+if __name__ == "__main__":
+    main()

samples/fizzbuzz.py.html

@@ -0,0 +1,8 @@
+<span class="keyword">def</span> <span class="function">main</span>():
+    <span class="keyword">for</span> <span class="variable">n</span> <span class="operator">in</span> <span class="function builtin">range</span>(1, 101):
+        <span class="variable">lut</span> <span class="operator">=</span> [<span class="variable">n</span>, <span class="string">"Fizz"</span>, <span class="string">"Buzz"</span>, <span class="string">"Fizz Buzz"</span>]
+        <span class="variable">idx</span> <span class="operator">=</span> (<span class="variable">n</span> <span class="operator">%</span> 3 <span class="operator">==</span> 0) <span class="operator">+</span> 2 <span class="operator">*</span> (<span class="variable">n</span> <span class="operator">%</span> 5 <span class="operator">==</span> 0)
+        <span class="function builtin">print</span>(<span class="variable">lut</span>[<span class="variable">idx</span>])
+
+<span class="keyword">if</span> <span class="variable">__name__</span> <span class="operator">==</span> <span class="string">"__main__"</span>:
+    <span class="function">main</span>()

samples/fizzbuzz.rs

@@ -0,0 +1,10 @@
+fn main() {
+    for n in 1..=100 {
+        match (n % 3, n % 5) {
+            (0, 0) => println!("Fizz buzz"),
+            (0, _) => println!("Fizz"),
+            (_, 0) => println!("Buzz"),
+            _ => println!("{}", n),
+        }
+    }
+}

samples/fizzbuzz.rs.html

@@ -0,0 +1,10 @@
+<span class="keyword">fn</span> <span class="function">main</span><span class="punctuation bracket">()</span> {
+    <span class="keyword">for</span> n <span class="keyword">in</span> <span class="constant">1</span>..=<span class="constant">100</span> {
+        <span class="keyword">match</span> <span class="punctuation bracket">(</span>n % <span class="constant">3</span>, n % <span class="constant">5</span><span class="punctuation bracket">)</span> {
+            <span class="punctuation bracket">(</span><span class="constant">0</span>, <span class="constant">0</span><span class="punctuation bracket">)</span> =&gt; <span class="function">println!</span><span class="punctuation bracket">(</span><span class="string">"Fizz buzz"</span><span class="punctuation bracket">)</span>,
+            <span class="punctuation bracket">(</span><span class="constant">0</span>, _<span class="punctuation bracket">)</span> =&gt; <span class="function">println!</span><span class="punctuation bracket">(</span><span class="string">"Fizz"</span><span class="punctuation bracket">)</span>,
+            <span class="punctuation bracket">(</span>_, <span class="constant">0</span><span class="punctuation bracket">)</span> =&gt; <span class="function">println!</span><span class="punctuation bracket">(</span><span class="string">"Buzz"</span><span class="punctuation bracket">)</span>,
+            _ =&gt; <span class="function">println!</span><span class="punctuation bracket">(</span><span class="string">"{}"</span>, n<span class="punctuation bracket">)</span>,
+        }
+    }
+}

src/languages.rs

@@ -1,27 +0,0 @@
-use tree_sitter_highlight::HighlightConfiguration;
-
-fn also_configure(mut config: HighlightConfiguration) -> (HighlightConfiguration, Vec<String>) {
-    let capture_names: Vec<_> = config
-        .query
-        .capture_names()
-        .iter()
-        .map(String::clone)
-        .collect();
-    config.configure(&capture_names);
-
-    (config, capture_names)
-}
-
-pub fn get_highlight_config(lang: &str) -> Option<(HighlightConfiguration, Vec<String>)> {
-    match lang {
-        "rust" => HighlightConfiguration::new(
-            tree_sitter_rust::language(),
-            tree_sitter_rust::HIGHLIGHT_QUERY,
-            "",
-            "",
-        )
-        .map(also_configure)
-        .ok(),
-        _ => None,
-    }
-}

src/languages/built_in.rs

@@ -0,0 +1,83 @@
+use tree_sitter::QueryError;
+use tree_sitter_highlight::HighlightConfiguration;
+
+use crate::register_language;
+
+pub static COMMON_HIGHLIGHT_NAMES: &[&str] = &[
+    "attribute",
+    "constant",
+    "function.builtin",
+    "function",
+    "keyword",
+    "operator",
+    "property",
+    "punctuation",
+    "punctuation.bracket",
+    "punctuation.delimiter",
+    "string",
+    "string.special",
+    "tag",
+    "type",
+    "type.builtin",
+    "variable",
+    "variable.builtin",
+    "variable.parameter",
+];
+
+pub fn register_builtin_languages() {
+    register_language("rust", rust_hl_factory);
+
+    register_language("javascript", javascript_hl_factory);
+    register_language("js", javascript_hl_factory);
+    register_language("jsx", jsx_hl_factory);
+
+    register_language("regex", regex_hl_factory);
+
+    register_language("python", tree_sitter_python);
+    register_language("py", tree_sitter_python);
+}
+
+pub fn rust_hl_factory() -> Result<HighlightConfiguration, QueryError> {
+    HighlightConfiguration::new(
+        tree_sitter_rust::language(),
+        tree_sitter_rust::HIGHLIGHT_QUERY,
+        "",
+        "",
+    )
+}
+
+pub fn javascript_hl_factory() -> Result<HighlightConfiguration, QueryError> {
+    HighlightConfiguration::new(
+        tree_sitter_javascript::language(),
+        tree_sitter_javascript::HIGHLIGHT_QUERY,
+        tree_sitter_javascript::INJECTION_QUERY,
+        tree_sitter_javascript::LOCALS_QUERY,
+    )
+}
+
+pub fn jsx_hl_factory() -> Result<HighlightConfiguration, QueryError> {
+    HighlightConfiguration::new(
+        tree_sitter_javascript::language(),
+        tree_sitter_javascript::JSX_HIGHLIGHT_QUERY,
+        tree_sitter_javascript::INJECTION_QUERY,
+        tree_sitter_javascript::LOCALS_QUERY,
+    )
+}
+
+pub fn regex_hl_factory() -> Result<HighlightConfiguration, QueryError> {
+    HighlightConfiguration::new(
+        tree_sitter_regex::language(),
+        tree_sitter_regex::HIGHLIGHTS_QUERY,
+        "",
+        "",
+    )
+}
+
+pub fn tree_sitter_python() -> Result<HighlightConfiguration, QueryError> {
+    HighlightConfiguration::new(
+        tree_sitter_python::language(),
+        tree_sitter_python::HIGHLIGHT_QUERY,
+        "",
+        "",
+    )
+}

src/languages/mod.rs

@@ -0,0 +1,32 @@
+use std::{collections::HashMap, sync::Mutex};
+
+use once_cell::sync::Lazy;
+use tree_sitter::QueryError;
+use tree_sitter_highlight::HighlightConfiguration;
+
+pub type HighlightConfigFactory = fn() -> Result<HighlightConfiguration, QueryError>;
+
+static LANGUAGES: Lazy<Mutex<HashMap<String, HighlightConfigFactory>>> =
+    Lazy::new(|| Mutex::new(HashMap::new()));
+
+#[cfg(feature = "built-in")]
+pub mod built_in;
+
+pub fn register_language(lang: impl Into<String>, factory: HighlightConfigFactory) {
+    let mut languages = LANGUAGES.lock().unwrap();
+    languages.insert(lang.into(), factory);
+}
+
+pub fn get_highlight_config(
+    lang: &str,
+    highlight_names: &[&str],
+) -> Option<HighlightConfiguration> {
+    let languages = LANGUAGES.lock().unwrap();
+    languages
+        .get(lang)
+        .and_then(|factory| factory().ok())
+        .map(|mut config| {
+            config.configure(highlight_names);
+            config
+        })
+}

src/lib.rs

@@ -3,16 +3,46 @@ use tree_sitter_highlight::{HighlightEvent, Highlighter};
 
 mod languages;
 
-pub fn highlight(source: &str) -> String {
-    let source = source.as_bytes();
+#[cfg(feature = "built-in")]
+pub use languages::built_in;
+pub use languages::register_language;
+
+fn get_hash_for_attrs(attrs: &[&str]) -> u64 {
+    use std::{
+        collections::hash_map::DefaultHasher,
+        hash::{Hash, Hasher},
+    };
+
+    let mut hasher = DefaultHasher::new();
+    attrs.hash(&mut hasher);
+    hasher.finish()
+}
+
+fn write_opening_tag(out: &mut String, attrs: &[&str]) {
+    out.push_str("<span class=\"");
+    for (i, attr) in attrs.iter().enumerate() {
+        if i != 0 {
+            out.push(' ');
+        }
+
+        out.push_str(&attr.replace('.', " "));
+    }
+    out.push_str("\">");
+}
+
+pub fn highlight(lang: &str, source: &str, highlight_names: &[&str]) -> String {
+    let highlight_config = match languages::get_highlight_config(lang, highlight_names) {
+        Some(conf) => conf,
+        None => return source.to_string(),
+    };
+
+    let source_bytes = source.as_bytes();
 
     let mut highlighter = Highlighter::new();
-    let (highlight_config, capture_names) =
-        languages::get_highlight_config("rust").expect("Could not get Rust language config");
-
     let items = highlighter
-        .highlight(&highlight_config, source, None, |lang| {
-            languages::get_highlight_config(lang).map(|x| /* ugh */ &*Box::leak(Box::new(x.0)))
+        .highlight(&highlight_config, source_bytes, None, |injected_lang| {
+            languages::get_highlight_config(injected_lang, highlight_names)
+                .map(|x| /* ugh */ &*Box::leak(Box::new(x)))
         })
         .unwrap();
 
@@ -21,58 +51,40 @@ pub fn highlight(source: &str) -> String {
 
     // Collapse adjacent identical attribute sets
     let mut last_attrs: u64 = 0;
-    let mut span_is_open = false;
+    let mut tag_is_open = false;
 
     for item in items {
         match item.unwrap() {
             HighlightEvent::Source { start, end } => {
-                let source_section_bytes = &source[start..end];
-                let source_section = String::from_utf8_lossy(source_section_bytes);
+                let source_section = &source[start..end];
 
-                let attr_hash = {
-                    use std::{
-                        collections::hash_map::DefaultHasher,
-                        hash::{Hash, Hasher},
-                    };
-
-                    let mut hasher = DefaultHasher::new();
-                    highlight_attrs.hash(&mut hasher);
-                    hasher.finish()
-                };
-
-                if last_attrs != attr_hash {
-                    if span_is_open {
-                        out.push_str("</span>");
-                        span_is_open = false;
-                    }
-
-                    if !highlight_attrs.is_empty() {
-                        out.push_str("<span class=\"");
-                        for (i, attr) in highlight_attrs.iter().enumerate() {
-                            if i != 0 {
-                                out.push(' ');
-                            }
-                            out.push_str(&attr.replace('.', " "));
-                        }
-                        out.push_str("\">");
-                        span_is_open = true;
-                        last_attrs = attr_hash;
-                    }
+                let attr_hash = get_hash_for_attrs(&highlight_attrs);
+                if last_attrs != attr_hash && tag_is_open {
+                    out.push_str("</span>");
+                    tag_is_open = false;
                 }
 
-                out.push_str(&encode_text(&source_section));
+                if !highlight_attrs.is_empty() && (!tag_is_open || last_attrs != attr_hash) {
+                    write_opening_tag(&mut out, &highlight_attrs);
+                    tag_is_open = true;
+                    last_attrs = attr_hash;
+                }
+
+                out.push_str(&encode_text(source_section));
             }
+
             HighlightEvent::HighlightStart(highlight) => {
-                let capture_name = &capture_names[highlight.0];
+                let capture_name = &highlight_names[highlight.0];
                 highlight_attrs.push(capture_name);
             }
+
             HighlightEvent::HighlightEnd => {
                 highlight_attrs.pop();
             }
         }
     }
 
-    if span_is_open {
+    if tag_is_open {
         out.push_str("</span>");
     }
 
@@ -80,18 +92,46 @@ pub fn highlight(source: &str) -> String {
 }
 
 #[cfg(test)]
+#[cfg(feature = "built-in")]
 mod tests {
-    use crate::highlight;
+    use crate::*;
 
     #[test]
     fn highlight_rust() {
-        println!(
-            "{}",
-            highlight(
-                r#"fn main() {
-    println!("Hello, world!");
-}"#
-            )
-        )
+        built_in::register_builtin_languages();
+
+        let source = include_str!("../samples/fizzbuzz.rs");
+        let expected_result = include_str!("../samples/fizzbuzz.rs.html");
+
+        assert_eq!(
+            highlight("rust", source, built_in::COMMON_HIGHLIGHT_NAMES),
+            expected_result
+        );
+    }
+
+    #[test]
+    fn highlight_js() {
+        built_in::register_builtin_languages();
+
+        let source = include_str!("../samples/fizzbuzz.js");
+        let expected_result = include_str!("../samples/fizzbuzz.js.html");
+
+        assert_eq!(
+            highlight("js", source, built_in::COMMON_HIGHLIGHT_NAMES),
+            expected_result
+        );
+    }
+
+    #[test]
+    fn highlight_python() {
+        built_in::register_builtin_languages();
+
+        let source = include_str!("../samples/fizzbuzz.py");
+        let expected_result = include_str!("../samples/fizzbuzz.py.html");
+
+        assert_eq!(
+            highlight("python", source, built_in::COMMON_HIGHLIGHT_NAMES),
+            expected_result
+        );
     }
 }