Construct Tree Sitter queries only once (#18)

* refactor: re-add main.rs and release-with-debug profile

This makes it easier to profile the program with a profiler (e.g. cargo
flamegraph), even if the Rust main won't be distributed.

* perf: create TS queries only once

Creating the queries isn't free, and when we're parsing lots of files in
directoy, it becomes a hotspot.

Cargo.toml

@@ -24,3 +24,7 @@ streaming-iterator = "0.1.9"
 
 [features]
 extension-module = ["pyo3/extension-module"]
+
+[profile.release-with-debug]
+inherits = "release"
+debug = true

src/input.rs

@@ -1,6 +1,7 @@
 use anyhow::Result;
 use std::fs;
 use std::path::PathBuf;
+use std::sync::LazyLock;
 use walkdir::WalkDir;
 
 #[derive(Debug, Clone, PartialEq)]
@@ -22,15 +23,34 @@ impl TryFrom<&PathBuf> for FileType {
     }
 }
 
+static TS_QUERY_PYTHON: LazyLock<tree_sitter::Query> = LazyLock::new(|| {
+    tree_sitter::Query::new(&tree_sitter_python::LANGUAGE.into(), "(comment) @comment")
+        .expect("Query must be valid")
+});
+
+static TS_QUERY_RUST: LazyLock<tree_sitter::Query> = LazyLock::new(|| {
+    tree_sitter::Query::new(
+        &tree_sitter_rust::LANGUAGE.into(),
+        "(line_comment) @comment
+(block_comment) @comment",
+    )
+    .expect("Query must be valid")
+});
+
+static TS_QUERY_JAVASCRIPT: LazyLock<tree_sitter::Query> = LazyLock::new(|| {
+    tree_sitter::Query::new(
+        &tree_sitter_javascript::LANGUAGE.into(),
+        "(comment) @comment",
+    )
+    .expect("Query must be valid")
+});
+
 impl FileType {
-    pub fn tree_sitter_query(&self) -> &'static str {
+    pub fn tree_sitter_query(&self) -> &'static tree_sitter::Query {
         match self {
-            FileType::Python => "(comment) @comment",
-            FileType::Rust => {
-                "(line_comment) @comment
-                (block_comment) @comment"
-            }
-            FileType::JavaScript => "(comment) @comment",
+            FileType::Python => &TS_QUERY_PYTHON,
+            FileType::Rust => &TS_QUERY_RUST,
+            FileType::JavaScript => &TS_QUERY_JAVASCRIPT,
         }
     }
 
@@ -88,17 +108,4 @@ mod tests {
         );
         assert!(determine_file_type(&PathBuf::from("test.txt")).is_err());
     }
-
-    #[test]
-    fn test_queries_are_valid() {
-        for file_type in [FileType::Python, FileType::Rust, FileType::JavaScript] {
-            let language = file_type.tree_sitter_language();
-            let query = file_type.tree_sitter_query();
-            assert!(
-                tree_sitter::Query::new(&language, query).is_ok(),
-                "Testing query from {:?}.",
-                file_type
-            );
-        }
-    }
 }

src/main.rs

@@ -0,0 +1,7 @@
+use anyhow::Result;
+
+fn main() -> Result<()> {
+    let args: Vec<String> = std::env::args().collect();
+    anot::cli::run(args)?;
+    Ok(())
+}

src/parser.rs

@@ -30,17 +30,16 @@ pub fn extract_annotations(
     let mut parser = tree_sitter::Parser::new();
     let language = file_type.tree_sitter_language();
     parser.set_language(&language)?;
-    let query = file_type.tree_sitter_query();
 
     // Parse the full source code
     let source = source_code.as_bytes();
     let tree = parser
         .parse(source, None)
         .ok_or_else(|| anyhow::anyhow!("Failed to parse source code"))?;
 
-    let query = tree_sitter::Query::new(&language, query)?;
+    let query = file_type.tree_sitter_query();
     let mut query_cursor = tree_sitter::QueryCursor::new();
-    let mut matches = query_cursor.matches(&query, tree.root_node(), source);
+    let mut matches = query_cursor.matches(query, tree.root_node(), source);
 
     let mut annotations = Vec::new();