IDX, EFT, ENTs Parsers (SEC/Edgar/Equity Research) (github.com)

The author released several parsing tools and updates for working with SEC/EDGAR and equity research data: an entity_classification.py (useful if you can observe regulatory data like RIAs from FINRA/SEC), a new ParserIDX that aims to unlock structured information across millions of companies, government bodies, and persons, and a revised ENTS module that can classify entities by filing form when that metadata is present (a ParserIDX use case). They also introduced parser_EFT.py, a first-draft tool to scan EFT records for subsidiary discovery by detecting EX21.1 exhibit appearances; initial output is available in EX21.json. Some classes/functions are placeholders and the author offers to finalize them if others want to run the tools on their devices. For the AI/ML community this matters because these parsers enable much richer entity resolution, corpus labeling, and relationship extraction from regulatory filings—foundational inputs for knowledge graphs, ownership networks, compliance automation, and equity-research models. Key technical points: ParserIDX targets broad-scale entity extraction across IDX-style records, ENTS now supports form-based classification, and parser_EFT focuses on EX21.1 exhibits to infer subsidiaries. The work is early-stage (sketch-level for EFT parsing) but promising for training downstream models or augmenting datasets once placeholders are replaced and data access is arranged.