Validates a text reference (tref) and returns structured metadata about it.

The response varies based on the node_type of the referenced text:

• JaggedArrayNode: Standard text nodes (e.g. biblical chapters/verses, Talmud pages). Returns depth, address_types, section_names, and positional indexes/labels.
• SchemaNode: Nodes with named children (e.g. a commentary with an introduction and main text). Returns children list.
• DictionaryEntryNode: Individual dictionary entries. Returns lexicon_name and headword, plus the JaggedArrayNode fields.
• DictionaryNode: Dictionary root nodes. Returns lexicon_name.
• SheetNode: User-created source sheets. Returns sheet_id.

If the node has a default child (a child node that represents the primary content), the response includes a default_child_node object with the child's node_type, node_index, and optionally depth, address_types and sectionNames (for JaggedArrayNode children) or lexicon_name (for DictionaryNode parents).

All types include navigation_refs with shortest_path_to_root (array of ancestor refs from immediate parent to book level) and first_available_section_ref. Segment-level refs also get prev_segment_ref/next_segment_ref, and section-level refs get prev_section_ref/next_section_ref.