Merge pull request #870 from posit-dev/feature/static-imports

Static diagnostics for `library()` and `require()` calls

Author: lionel-

crates/ark/src/lsp/diagnostics.rs

@@ -309,10 +309,12 @@ mod tests {
     use crate::lsp::diagnostics::DiagnosticContext;
     use crate::lsp::diagnostics_syntax::syntax_diagnostics;
     use crate::lsp::documents::Document;
+    use crate::lsp::inputs::library::Library;
 
     fn text_diagnostics(text: &str) -> Vec<Diagnostic> {
         let document = Document::new(text, None);
-        let context = DiagnosticContext::new(&document.contents);
+        let library = Library::default();
+        let context = DiagnosticContext::new(&document.contents, &library);
         let diagnostics = syntax_diagnostics(document.ast.root_node(), &context).unwrap();
         diagnostics
     }

crates/ark/src/lsp/diagnostics_syntax.rs

@@ -19,6 +19,7 @@ use stdext::unwrap;
 use stdext::unwrap::IntoResult;
 use tower_lsp::lsp_types::Range;
 use tree_sitter::Node;
+use tree_sitter::Query;
 use walkdir::DirEntry;
 use walkdir::WalkDir;
 
@@ -29,7 +30,7 @@ use crate::lsp::traits::rope::RopeExt;
 use crate::treesitter::BinaryOperatorType;
 use crate::treesitter::NodeType;
 use crate::treesitter::NodeTypeExt;
-use crate::treesitter::TSQuery;
+use crate::treesitter::TsQuery;
 
 #[derive(Clone, Debug)]
 pub enum IndexEntryData {
@@ -347,23 +348,27 @@ fn index_r6_class_methods(
     entries: &mut Vec<IndexEntry>,
 ) -> anyhow::Result<()> {
     // Tree-sitter query to match individual methods in R6Class public/private lists
-    let query_str = r#"
-    (argument
-      name: (identifier) @access
-      value: (call
-        function: (identifier) @_list_fn
-        arguments: (arguments
-          (argument
-            name: (identifier) @method_name
-            value: (function_definition) @method_fn
-          )
-        )
-      )
-      (#match? @access "public|private")
-      (#eq? @_list_fn "list")
-    )
-    "#;
-    let mut ts_query = TSQuery::new(query_str)?;
+    static R6_METHODS_QUERY: LazyLock<Query> = LazyLock::new(|| {
+        let query_str = r#"
+            (argument
+                name: (identifier) @access
+                value: (call
+                    function: (identifier) @_list_fn
+                    arguments: (arguments
+                        (argument
+                            name: (identifier) @method_name
+                            value: (function_definition) @method_fn
+                        )
+                    )
+                )
+                (#match? @access "public|private")
+                (#eq? @_list_fn "list")
+            )
+        "#;
+        let language = &tree_sitter_r::LANGUAGE.into();
+        Query::new(language, query_str).expect("Failed to compile R6 methods query")
+    });
+    let mut ts_query = TsQuery::from_query(&*R6_METHODS_QUERY);
 
     // We'll switch from Rope to String in the near future so let's not
     // worry about this conversion now

crates/ark/src/lsp/indexer.rs

@@ -0,0 +1,147 @@
+//
+// library.rs
+//
+// Copyright (C) 2025 by Posit Software, PBC
+//
+
+use std::collections::HashMap;
+use std::path::PathBuf;
+use std::sync::Arc;
+use std::sync::RwLock;
+
+use super::package::Package;
+use crate::lsp;
+
+/// Lazily manages a list of known R packages by name
+#[derive(Default, Clone, Debug)]
+pub struct Library {
+    /// Paths to library directories, i.e. what `base::libPaths()` returns.
+    pub library_paths: Arc<Vec<PathBuf>>,
+
+    packages: Arc<RwLock<HashMap<String, Option<Arc<Package>>>>>,
+}
+
+impl Library {
+    pub fn new(library_paths: Vec<PathBuf>) -> Self {
+        Self {
+            packages: Arc::new(RwLock::new(HashMap::new())),
+            library_paths: Arc::new(library_paths),
+        }
+    }
+
+    /// Get a package by name, loading and caching it if necessary.
+    /// Returns `None` if the package can't be found or loaded.
+    pub fn get(&self, name: &str) -> Option<Arc<Package>> {
+        // Try to get from cache first (could be `None` if we already tried to
+        // load a non-existent or broken package)
+        if let Some(entry) = self.packages.read().unwrap().get(name) {
+            return entry.clone();
+        }
+
+        // Not cached, try to load
+        let pkg = match self.load_package(name) {
+            Ok(Some(pkg)) => Some(Arc::new(pkg)),
+            Ok(None) => None,
+            Err(err) => {
+                lsp::log_error!("Can't load R package: {err:?}");
+                None
+            },
+        };
+
+        self.packages
+            .write()
+            .unwrap()
+            .insert(name.to_string(), pkg.clone());
+
+        pkg
+    }
+
+    /// Insert a package in the library for testing purposes.
+    #[cfg(test)]
+    pub fn insert(self, name: &str, package: Package) -> Self {
+        self.packages
+            .write()
+            .unwrap()
+            .insert(name.to_string(), Some(Arc::new(package)));
+        self
+    }
+
+    fn load_package(&self, name: &str) -> anyhow::Result<Option<Package>> {
+        for lib_path in self.library_paths.iter() {
+            match Package::load(&lib_path, name)? {
+                Some(pkg) => return Ok(Some(pkg)),
+                None => (),
+            }
+        }
+
+        Ok(None)
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use std::fs::File;
+    use std::fs::{self};
+    use std::io::Write;
+
+    use tempfile::TempDir;
+
+    use super::*;
+
+    // Helper to create a temporary package directory with DESCRIPTION and NAMESPACE
+    fn create_temp_package(
+        pkg_name: &str,
+        description: &str,
+        namespace: &str,
+    ) -> (TempDir, PathBuf) {
+        let temp_dir = TempDir::new().unwrap();
+        let pkg_dir = temp_dir.path().join(pkg_name);
+        fs::create_dir(&pkg_dir).unwrap();
+
+        let desc_path = pkg_dir.join("DESCRIPTION");
+        let mut desc_file = File::create(&desc_path).unwrap();
+        desc_file.write_all(description.as_bytes()).unwrap();
+
+        let ns_path = pkg_dir.join("NAMESPACE");
+        let mut ns_file = File::create(&ns_path).unwrap();
+        ns_file.write_all(namespace.as_bytes()).unwrap();
+
+        (temp_dir, pkg_dir)
+    }
+
+    #[test]
+    fn test_load_and_cache_package() {
+        let pkg_name = "mypkg";
+        let description = r#"
+Package: mypkg
+Version: 1.0
+        "#;
+        let namespace = r#"
+export(foo)
+export(bar)
+importFrom(pkg, baz)
+        "#;
+
+        let (temp_dir, _pkg_dir) = create_temp_package(pkg_name, description, namespace);
+
+        // Library should point to the temp_dir as its only library path
+        let lib = Library::new(vec![temp_dir.path().to_path_buf()]);
+
+        // First access loads from disk
+        let pkg = lib.get(pkg_name).unwrap();
+        assert_eq!(pkg.description.name, "mypkg");
+
+        // Second access uses cache (note that we aren't testing that we are
+        // indeed caching, just exercising the cache code path)
+        assert!(lib.get(pkg_name).is_some());
+
+        // Negative cache: missing package
+        assert!(lib.get("notapkg").is_none());
+        // Now cached as absent
+        assert!(lib.get("notapkg").is_none());
+
+        // Namespace is parsed
+        assert_eq!(pkg.namespace.exports, vec!["bar", "foo"]);
+        assert_eq!(pkg.namespace.imports, vec!["baz"]);
+    }
+}

crates/ark/src/lsp/inputs/library.rs

@@ -0,0 +1,12 @@
+//
+// mod.rs
+//
+// Copyright (C) 2025 by Posit Software, PBC
+//
+//
+
+pub mod library;
+pub mod package;
+pub mod package_description;
+pub mod package_namespace;
+pub mod source_root;

crates/ark/src/lsp/inputs/mod.rs

@@ -0,0 +1,64 @@
+//
+// package.rs
+//
+// Copyright (C) 2025 by Posit Software, PBC
+//
+//
+
+use std::fs;
+use std::path::PathBuf;
+
+use crate::lsp::inputs::package_description::Description;
+use crate::lsp::inputs::package_namespace::Namespace;
+
+/// Represents an R package and its metadata relevant for static analysis.
+#[derive(Clone, Debug)]
+pub struct Package {
+    /// Path to the directory that contains `DESCRIPTION``. Can
+    /// be an installed package or a package source.
+    pub path: PathBuf,
+
+    pub description: Description,
+    pub namespace: Namespace,
+}
+
+impl Package {
+    /// Attempts to load a package from the given path and name.
+    pub fn load(lib_path: &std::path::Path, name: &str) -> anyhow::Result<Option<Self>> {
+        let package_path = lib_path.join(name);
+
+        let description_path = package_path.join("DESCRIPTION");
+        let namespace_path = package_path.join("NAMESPACE");
+
+        // Only consider libraries that have a folder named after the
+        // requested package and that contains a description file
+        if !description_path.is_file() {
+            return Ok(None);
+        }
+
+        // This fails if there is no `Package` field, so we're never loading
+        // folders like bookdown projects as package
+        let description_contents = fs::read_to_string(&description_path)?;
+        let description = Description::parse(&description_contents)?;
+
+        if description.name != name {
+            return Err(anyhow::anyhow!(
+                "`Package` field in `DESCRIPTION` doesn't match folder name '{name}'"
+            ));
+        }
+
+        let namespace = if namespace_path.is_file() {
+            let namespace_contents = fs::read_to_string(&namespace_path)?;
+            Namespace::parse(&namespace_contents)?
+        } else {
+            tracing::info!("Package `{name}` doesn't contain a NAMESPACE file, using defaults");
+            Namespace::default()
+        };
+
+        Ok(Some(Package {
+            path: package_path,
+            description,
+            namespace,
+        }))
+    }
+}