Distinct files in the full Software Heritage archive (2026-06-04 export), dated by first appearance, following Desmazières, Di Cosmo, Lorentz (MSR 2025). Case-aggregated. citation.
5.3M
total files
2.9M
first seen 2021–2025 (54.9%)
2021
median year (half its files since)
Case variants in archive: .h5 (5.3M), .H5 (982) — aggregated above.
Attribution status
This extension is attributed to hdf — every claim across all sources (including secondary ones) points to a single language.
Authoritative sources:✗ Linguist · no claim✗ Pygments · no claim△ Wikidata · secondary✗ Wikipedia · no claim
File formats, image formats, audio codecs and other non-PL
entities that claim .h5 on Wikidata
(property P1195) or in the Wikipedia infobox.
Source: data/derived/external_extension_index.csv.
Click Use as new PL on any row to open the
Add-PL form pre-filled with that entry's name, Wikipedia/Wikidata
URL, and this extension — one submission creates the PL + claim.
Real archived programs with this extension, byte-verified against the SWH archive, grouped by predicted language and by how the prediction was made. Useful for deciding what this extension actually is when the attribution is uncertain.