US 12,737,390 B1
Search labeling system for enhanced content indexing and retrieval
Matthew Weidner, San Francisco, CA (US); Stephan Boyer, San Francisco, CA (US); Michael Kuo, Long Island City, NY (US); Benjamin Schley Hughes, Brooklyn, NY (US); Fabricio Pontes Harsich, San Francisco, CA (US); and Emma Yuchen Guo, San Francisco, CA (US)
Assigned to Notion Labs, Inc., San Francisco, CA (US)
Filed by Notion Labs, Inc., San Francisco, CA (US)
Filed on Feb. 23, 2026, as Appl. No. 19/547,620.
Claims priority of provisional application 63/891,938, filed on Oct. 1, 2025.
Int. Cl. G06F 16/31 (2019.01); G06F 16/334 (2025.01)
CPC G06F 16/322 (2019.01) [G06F 16/334 (2019.01)] 19 Claims
OG exemplary drawing
 
16. A system comprising:
at least one hardware processor; and
at least one non-transitory memory storing instructions, which, when executed by the at least one hardware processor, cause the system to:
create a block within a digital workspace that includes multiple blocks,
wherein the block is a discrete unit of content including text that is a conflict-free replicated data type (CRDT) instance that is assigned a text instance identifier (ID);
store a single text slice tree in a database record linked to the block,
wherein the text slice tree has a hierarchical structure in which each node represents a text slice corresponding to a segment of the text, and
wherein a database table stores the text instance ID for the text and a search label for each text slice including a parent search label for a root node, each search label including a string identifier;
in response to a text editing operation to split the text of the block:
segment the text into two contiguous text slices, represented as a first child node and a second child node in the text slice tree;
propagate the parent search label to both the first and second child nodes;
assign to the first child node a first child search label that comprises the parent search label as a prefix appended with a first character;
assign to the second child node a second child search label that comprises the parent search label as a prefix appended with a second character, the second character being different from the first character; and
store the first and second child search labels in the database table,
wherein a combination of the text instance ID and a particular search label for a particular node of the text slice tree uniquely identifies a corresponding particular text slice in the digital workspace, and
wherein associations between search labels and their respective text slices are preserved throughout text editing operations performed on the text of the block.