tree-sitter-core

Tree-sitter generates a Concrete Syntax Tree (CST), not a true AST, because it retains anonymous nodes like punctuation alongside named nodes. This preserves the complete source structure for accurate syntax representation and debugging.

To identify named vs unnamed nodes in tree-sitter, you inspect the node's type: named nodes have descriptive identifiers (e.g., `function_declaration`), while unnamed nodes represent punctuation or keywords. This distinction is crucial for debugging structural patterns.

ast-grep maps to tree-sitter by matching named nodes and using fields to define parent-child relationships. This mapping clarifies how ast-grep patterns translate into underlying tree-sitter queries for precise syntax tree navigation.

ast-grep patterns often fail to match unnamed nodes because ast-grep primarily targets named tree-sitter nodes. Unnamed nodes, such as punctuation or keywords, are typically ignored in pattern matching unless explicitly referenced in the query syntax.

Yes, you can use tree-sitter to inspect parse trees across multiple programming languages. Its generalized parsing model standardizes node classification into named and anonymous types, enabling consistent syntax tree analysis and debugging regardless of the language grammar.

In tree-sitter, kinds refer to the specific node type (e.g., `identifier`), while fields define the structural role of a child node within its parent (e.g., `left` or `right` in a binary operation). Distinguishing them is essential for accurate tree navigation.