[ty] AST garbage collection (#18482)

## Summary

Garbage collect ASTs once we are done checking a given file. Queries
with a cross-file dependency on the AST will reparse the file on demand.
This reduces ty's peak memory usage by ~20-30%.

The primary change of this PR is adding a `node_index` field to every
AST node, that is assigned by the parser. `ParsedModule` can use this to
create a flat index of AST nodes any time the file is parsed (or
reparsed). This allows `AstNodeRef` to simply index into the current
instance of the `ParsedModule`, instead of storing a pointer directly.

The indices are somewhat hackily (using an atomic integer) assigned by
the `parsed_module` query instead of by the parser directly. Assigning
the indices in source-order in the (recursive) parser turns out to be
difficult, and collecting the nodes during semantic indexing is
impossible as `SemanticIndex` does not hold onto a specific
`ParsedModuleRef`, which the pointers in the flat AST are tied to. This
means that we have to do an extra AST traversal to assign and collect
the nodes into a flat index, but the small performance impact (~3% on
cold runs) seems worth it for the memory savings.

Part of https://github.com/astral-sh/ty/issues/214.

crates/ruff_python_formatter/src/comments/debug.rs

@@ -184,7 +184,7 @@ mod tests {
     use insta::assert_debug_snapshot;
 
     use ruff_formatter::SourceCode;
-    use ruff_python_ast::AnyNodeRef;
+    use ruff_python_ast::{AnyNodeRef, AtomicNodeIndex};
     use ruff_python_ast::{StmtBreak, StmtContinue};
     use ruff_python_trivia::{CommentLinePosition, CommentRanges};
     use ruff_text_size::{TextRange, TextSize};
@@ -196,10 +196,12 @@ mod tests {
     fn debug() {
         let continue_statement = StmtContinue {
             range: TextRange::new(TextSize::new(18), TextSize::new(26)),
+            node_index: AtomicNodeIndex::dummy(),
         };
 
         let break_statement = StmtBreak {
             range: TextRange::new(TextSize::new(55), TextSize::new(60)),
+            node_index: AtomicNodeIndex::dummy(),
         };
 
         let source = r"# leading comment

crates/ruff_python_formatter/src/comments/node_key.rs

@@ -53,6 +53,7 @@ impl<'a> From<AnyNodeRef<'a>> for NodeRefEqualityKey<'a> {
 mod tests {
     use crate::comments::node_key::NodeRefEqualityKey;
     use ruff_python_ast::AnyNodeRef;
+    use ruff_python_ast::AtomicNodeIndex;
     use ruff_python_ast::StmtContinue;
     use ruff_text_size::TextRange;
     use std::collections::hash_map::DefaultHasher;
@@ -68,6 +69,7 @@ mod tests {
     fn equality() {
         let continue_statement = StmtContinue {
             range: TextRange::default(),
+            node_index: AtomicNodeIndex::dummy(),
         };
 
         let ref_a = NodeRefEqualityKey::from_ref(AnyNodeRef::from(&continue_statement));
@@ -81,6 +83,7 @@ mod tests {
     fn inequality() {
         let continue_statement = StmtContinue {
             range: TextRange::default(),
+            node_index: AtomicNodeIndex::dummy(),
         };
 
         let boxed = Box::new(continue_statement.clone());

crates/ruff_python_formatter/src/comments/placement.rs

@@ -1067,6 +1067,7 @@ fn handle_slice_comments<'a>(
 ) -> CommentPlacement<'a> {
     let ast::ExprSlice {
         range: _,
+        node_index: _,
         lower,
         upper,
         step,
@@ -1450,6 +1451,7 @@ fn handle_expr_if_comment<'a>(
 ) -> CommentPlacement<'a> {
     let ast::ExprIf {
         range: _,
+        node_index: _,
         test,
         body,
         orelse,