TenK.business returns None when filing uses combined "Items 1 and 2" heading

[FlorinAndrei]

Issue Type

• Data Quality Issue (incorrect financial values, missing data)
• XBRL Parsing Issue (statement rendering, concept mapping)
• Filing Access Issue (download failures, attachment problems)
• Performance Issue (slow operations, memory problems)
• Compatibility Issue (platform/dependency conflicts)
• Documentation Issue (unclear instructions, missing examples)

Environment

EdgarTools Version: 5.23.3
Python Version: 3.13.12
Operating System: Ubuntu 24.04

Bug Description

What you expected to happen:
I expected to extract Item 1 from the 10-K of VNOM.

What actually happened:

The issue: when a 10-K uses a combined heading like "Items 1 and 2. Business and Properties" (instead of separate Item 1 and Item 2 sections), TenK.business returns None. The new parser correctly identifies the section as part_i_items_1_and_2._business_and_properties, but the business property (which maps to self['Item 1']) doesn't resolve it. The legacy fallback also fails.

Concrete example: VNOM (Viper Energy), accession 0002074176-26-000010. The 10-K has "ITEMS 1 and 2. BUSINESS AND PROPERTIES" at the Part I level. filing.obj().sections shows part_i_items_1_and_2._business_and_properties with content, but filing.obj().business returns None.

Error message (if any):

2026-03-16 06:50:34,824 INFO httpxthrottlecache.ratelimiter: Making HTTP Request <Request('GET', 'https://www.sec.gov/Archives/edgar/data/2074176/000207417626000010/vnom-20251231.htm')>
2026-03-16 06:50:34,841 INFO httpx: HTTP Request: GET https://www.sec.gov/Archives/edgar/data/2074176/000207417626000010/vnom-20251231.htm "HTTP/1.1 200 OK"
2026-03-16 06:50:36,469 INFO edgar.documents.extractors.toc_section_detector: TOC detection found 28 sections
2026-03-16 06:50:36,470 INFO edgar.documents.extractors.hybrid_section_detector: TOC detection successful: 28 sections found
2026-03-16 06:50:36,470 WARNING edgar.core: TenK falling back to legacy parser for 'Item 1' (filing: 0002074176-26-000010). New parser sections available: ['part_ii_item_9', 'part_ii_item_9a', 'part_iii_item_10', 'part_iii_item_11', 'part_iii_item_12', 'part_iii_item_13', 'part_ii_item_9b', 'part_ii_item_9c', 'part_iv_item_15', 'part_iv_item_16', 'part_i_items_1_and_2._business_and_properties', 'part_iv_items_1_and_2._business_and_properties', 'part_i_item_1a', 'part_ii_item_1a', 'part_iv_item_1a', 'part_i_item_1b', 'part_i_item_1c', 'part_iv_item_1c', 'part_i_item_3', 'part_ii_item_5', 'part_iv_item_5', 'part_ii_item_6', 'part_ii_item_7', 'part_iv_item_7', 'part_ii_item_7a', 'part_iv_item_7a', 'part_ii_item_8', 'part_iv_item_8']. This fallback will be removed in v6.0.

Reproduction

Company/Ticker: VNOM
Time Period: most recent
Relevant Forms: 10-K

Minimal code to reproduce:

from edgar import Company, set_identity

set_identity("Your Name your.email@example.com")
company = Company("VNOM")
filing = company.latest("10-K")
tenk = filing.obj()
print(tenk.business)  # None
print([s for s in str(tenk.sections) if 'items_1' in str(s)])  # part_i_items_1_and_2._business_and_properties exists

Additional Context

Add any other context about the problem here. Screenshots, related issues, etc.

---

This issue will be handled using EdgarTools' systematic issue resolution workflow. A reproduction test will be created to verify the fix.

[dgunning]

Thank you for the clear reproduction case — the accession number made this immediately verifiable.

Root cause confirmed. The new HTMLParser correctly identifies the section as part_i_items_1_and_2._business_and_properties, but the TenK.__getitem__ lookup builds the key part_i_item_1 and never checks the combined-items variant. All six lookup priority levels miss it, so the method falls through to the legacy ChunkedDocument fallback, which also cannot find the section under that heading format.

Fix is straightforward. We need to add a combined-items key scan to the lookup chain in ten_k.py. This is about five lines of code and touches nothing else. The same fix will handle ten_k['Item 2'] (Properties) for filings that combine Items 1 and 2.

Release timeline. We have a point release (v5.23.4) shipping today for a separate data accuracy issue. We are bundling this fix into that same release. You should see the fix in v5.23.4 today.

Scope note. Combined "Items 1 and 2. Business and Properties" headings appear in multiple energy, MLP, and REIT filers. If you know of other companies using this pattern, please share them so we can add regression coverage beyond VNOM.

Tracking: edgartools-xp44

[FlorinAndrei]

AFAICT I only see this issue with two symbols:

• DVN — accession 0001193125-26-056485
• VNOM — accession 0002074176-26-000010

Please double-check this information, I'm only working from error logs here.

Thanks!

[dgunning]

Fixed in v5.23.4. TenK.getitem now resolves combined "Items 1 and 2" headings for both item numbers. Verified with VNOM and DVN 10-K filings.