yeast: Support separate output node types in extractor generator

Language and LanguageSpec gain optional output_node_types field.
When set, the generator produces dbscheme/QL from the output types
and the extractor validates TRAP against them.

All existing extractors pass None (no behavior change).
Ruby extract() calls gain vec![] for the new rules parameter.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Author: tausbn

ql/extractor/src/generator.rs

@@ -21,18 +21,22 @@ pub fn run(options: Options) -> std::io::Result<()> {
         Language {
             name: "QL".to_owned(),
             node_types: tree_sitter_ql::NODE_TYPES,
+            output_node_types: None,
         },
         Language {
             name: "Dbscheme".to_owned(),
             node_types: tree_sitter_ql_dbscheme::NODE_TYPES,
+            output_node_types: None,
         },
         Language {
             name: "Blame".to_owned(),
             node_types: tree_sitter_blame::NODE_TYPES,
+            output_node_types: None,
         },
         Language {
             name: "JSON".to_owned(),
             node_types: tree_sitter_json::NODE_TYPES,
+            output_node_types: None,
         },
     ];
 

ruby/extractor/src/generator.rs

@@ -21,10 +21,12 @@ pub fn run(options: Options) -> std::io::Result<()> {
         Language {
             name: "Ruby".to_owned(),
             node_types: tree_sitter_ruby::NODE_TYPES,
+            output_node_types: None,
         },
         Language {
             name: "Erb".to_owned(),
             node_types: tree_sitter_embedded_template::NODE_TYPES,
+            output_node_types: None,
         },
     ];
 

shared/tree-sitter-extractor/src/generator/language.rs

@@ -1,4 +1,8 @@
 pub struct Language {
     pub name: String,
     pub node_types: &'static str,
+    /// If set, the generator uses these node types for the dbscheme/QL library
+    /// instead of `node_types`. This is useful when desugaring transforms produce
+    /// an AST whose shape differs from the tree-sitter grammar.
+    pub output_node_types: Option<&'static str>,
 }

shared/tree-sitter-extractor/src/generator/mod.rs

@@ -68,7 +68,8 @@ pub fn generate(
         let token_name = format!("{}_token", &prefix);
         let tokeninfo_name = format!("{}_tokeninfo", &prefix);
         let reserved_word_name = format!("{}_reserved_word", &prefix);
-        let nodes = node_types::read_node_types_str(&prefix, language.node_types)?;
+        let effective_node_types = language.output_node_types.unwrap_or(language.node_types);
+        let nodes = node_types::read_node_types_str(&prefix, effective_node_types)?;
         let (dbscheme_entries, mut ast_node_members, token_kinds) = convert_nodes(&nodes);
         ast_node_members.insert(&token_name);
         writeln!(&mut dbscheme_writer, "/*- {} dbscheme -*/", language.name)?;