Changelog

v01.0.3 - modified item mapping dict regex to be more robust. - modified Monitor().monitor_submissions() to have a start_date parameter.

v01.0.2 2025-02-06 - modified .parse() to return dictionary in previous package format.

v01.0.0 2025-02-04 - reworked parsing into doc2dict. - reworked sgml parsing into secsgml - reworked datasetbuilder into txt2dataset - fixed some downloader issues - made all xmls parseable - generalized documents iterablity - may break some work flows. Sorry - modified Downloader() to be more specific. downloading 10-K would also download 10-K/A.

v.433 (2025-01-21) - added deduplication to Premium Downloader

v.428 (2025-01-03) - added option to set download source via Config class - added option to search by document type and by containing text to Portfolio class - added download option to Portfolio class. Now you don’t need to import Downloader class to download submissions

v0.418 (2024-12-17) - improved imports

v0.417 (2024-12-17) - Fixed bug preventing installation on linux/macOS.

Related update

• SEC Library cache resets every month. Unfortunately, some files were accessed before uploaded, causing a cache miss. This will be fixed on the next cache reset.

v0.400 (2024-12-16)

• Major update. Reworked/removed some functions to improve performance and usability.

• If functionality removed was critical to your workflow, please post an issue on GitHub. Apologies.

• Reworked downloader to use .sgml submission files. This grabs the full submission, including attachments and metadata.

• Added premium downloader for faster downloads. This is a paid feature.

• Changed to Submission/Document terminology to be more inline with SEC taxonomy.

• specifying document type is no longer needed for parsing, as it is now inferred from the submission.

• reworked downloader.watch to monitor.monitor, and made callback functions better.

• reworked updating package data to packageupdater class

• added fast sgml parsing using cython

• this is an early release, so please report any issues on GitHub.

v0.382 (2024-11-19)

• Fixed issue where SEC recorded primary doc URL as 0001.extension, while the true location was acc no - 0001. Example: https://www.sec.gov/Archives/edgar/data/1102262/000110226201000003/0001102262-01-000003-0001.txt.

v0.379 (2024-11-18)

• Added metadata.

• Moved return URLs to metadata.

v0.377 (2024-11-01)

• Fixed SC 13G Item 10 being detected as Item 1.

v0.376 (2024-11-01)

• Added parsing and iterable objects for forms: 3, 13F-HR, NPORT-P, SC 13D, SC 13G, 10-Q, 10-K, 8-K, D.

v0.374 (2024-10-29)

• Added filing.write_json and filing.write_csv.

v0.373 (2024-10-29)

• Made parsed filing structure for 8-K, 10-K, 10-Q more intuitive.

v0.372 (2024-10-29)

• Reduced package import time to ~400 milliseconds. Further optimization planned for dependency cleanup.

v0.368 (2024-10-28)

• Improved parsing robustness for 10-K, 10-Q, and 8-K by centralizing helper scripts for loading file content and cleaning titles.

v0.364 (2024-10-28)

• Added parsing support for 10-K and 10-Q.

v0.363 (2024-10-26)

• Added dataset_builder.

• Enhanced 8-K parsing robustness.

v0.357 (2024-10-24)

• Improved access to package data.

v0.356 (2024-10-23)

• Added parsing support for 13F-HR information table and 8-K.

v0.355 (2024-10-23)

• Added setuptools to package for handling edge cases.

v0.352 (2024-10-21)

• Switched 10-K dataset download source to Dropbox from Zenodo.

v0.351 (2024-10-18)

• Added download options for attachments by file type and item type.

• Introduced up-to-date dataset for 13F-HR information tables.

v0.350 (2024-10-17)

• Added bulk download functionality for 10-K.

v0.343 (2024-10-16)

• Added bulk download functionality for 10-Q.

v0.342 (2024-10-16)

• Introduced callback function option for downloader.watch().

v0.341 (2024-10-15)

• Added company metadata datasets, including SIC codes, former names, and more.

v0.337 (2024-10-14)

• Added filtering options by SICs and items to downloader.

• Included FTD dataset in download_dataset.

v0.335 (2024-10-13)

• Added prefill option for MuleBot server.

v0.334 (2024-10-13)

• Added links to GitHub and website for chatbot.

v0.333 (2024-10-13)

• Simplified MuleBot server UI.

• Refactored MuleBot server into multiple modules.

v0.332 (2024-10-05)

• Modified table parser to output parsed tables in list format.

v0.330 (2024-10-03)

• Improved downloader robustness.

• Introduced set_limiter for precise control.

• Added dataset of all 10-Ks since 2001.

v0.323 (2024-09-27)

• Added MuleBot.

• Reworked Filing Viewer.

v0.314 (2024-09-26)

• Added TableParser.

v0.312 (2024-09-20)

• Introduced download_company_concepts.

v0.311 (2024-09-19)

• Added basic MuleBot tool calling and interface.

v0.302 (2024-09-18)

• Re-added output directory option to download functionality.

v0.301 (2024-09-18)

• Fixed Jupyter Notebook package data issue.

v0.29 (2024-09-18)

• Major overhaul: - Removed need to download or construct indices. - Expanded scope to all SEC filings since 2001, including companies without tickers and individuals. - Moved Indexer().watch() to downloader. - Temporarily removed filtering by company name due to exact matching issues.

v0.26 (2024-09-16)

• Introduced indexer.watch(interval, cik, form) for EDGAR monitoring.

v0.25 (2024-09-16)

• Added human_readable option to download and download_using_api.

Earlier Updates

• 2024-09-15: Fixed issue where filings download would overwrite each other.

• 2024-09-14: Added parser API support.

• 2024-09-13: Introduced download_datasets and index download options.

• 2024-09-09: Added download_using_api (no indices required).

• 2024-09-08: Integrated with datamule’s SEC Router API.

• 2024-09-07: Simplified indices approach, switched to Polar for faster index loading.