kglite-datasets
Fetch-build-cache dataset loaders for kglite —
pull a public registry, build it into a queryable kglite KnowledgeGraph, cache
the result. Extracted from kglite::datasets into a standalone crate + wheel so
the kglite engine links zero network code.
Bundled sources:
- SEC EDGAR — quarterly index, bulk submissions, per-form fetchers, XBRL company facts.
- Sodir — Norwegian Offshore Directorate registry.
- Wikidata — dump download / cache orchestration (heavy ingestion stays engine-side).
Install
pip install kglite-datasets # ships the loaders + Python interface; pulls kglite
The single wheel bundles the Rust loaders and the Python interface; the graph
engine is reused from kglite (a runtime dependency), not forked.
Usage
Each loader fetches its source, builds it into a kglite KnowledgeGraph, and
caches the result under a workdir — returning a graph ready to query:
# SEC EDGAR — name a form, a company, a span:
from kglite_datasets.sec import SEC
g = SEC.fetch(workdir, "13F-HR", "TSLA", years=2, user_agent="Name email@dom")
# Sodir (Norwegian Offshore Directorate):
from kglite_datasets import sodir
g = sodir.open(workdir)
# Wikidata truthy dump (cache-managed download + N-Triples build):
from kglite_datasets import wikidata
g = wikidata.open(workdir)
# Each call above returns a KnowledgeGraph — query the one you built:
g.cypher("MATCH (n) RETURN labels(n) AS labels, count(*) AS nodes ORDER BY nodes DESC LIMIT 5")
Importing sec never drags in sodir/wikidata's optional stack (lazy
submodules), so the SEC path stays light.
Rust
Rust embedders use the crate-root modules directly (they mirror kglite's former
api::datasets::<loader> surface): kglite_datasets::{sec, sodir, wikidata}.
Status
Alpha — extraction complete. Followed the codingest handover playbook:
copy → prove parity → freeze goldens + bench baseline (while kglite still had its
in-tree copy) → kglite removed its loaders (2026-07-16). With the in-tree copy
gone, the frozen goldens (tests/goldens/) and the frozen bench baseline
(benchmarks/baseline.json) are the sole oracles carrying that verified-correct
authority forward.
Development
make gate # CI-equivalent local gate: lint · build · test · determinism · bench-smoke
make lint # cargo fmt --check + clippy -D warnings (+ ruff if installed)
make develop # maturin develop — build the extension into the active venv
make pytest # offline Python suite (live-API tests behind the `live` marker)
The Rust workspace: crates/kglite-datasets (engine-free loaders) +
crates/kglite-datasets-py (PyO3 wrapper → the wheel). Loaders fetch → parse →
emit CSV/blueprint/dump; the graph build reuses kglite's from_blueprint /
load_ntriples. All tests are offline (recorded HTTP fixtures); live-API
smokes are opt-in.
Links
- Documentation: https://kglite-datasets.readthedocs.io
- PyPI: https://pypi.org/project/kglite-datasets/ (
pip install kglite-datasets) - crates.io: https://crates.io/crates/kglite-datasets (
cargo add kglite-datasets) - kglite engine: https://github.com/kkollsga/kglite
Badges and links point at the initial 0.1.0 release; there are no prior
published versions.
License
MIT © Kristian dF Kollsgård
1 maintainer
Metadata
1 maintainer
Release files for kglite-datasets 0.1.28
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| kglite_datasets-0.1.28.tar.gz | 300.6 kB | Details |
Built distributions (wheels)
| File | Reset | |||
|---|---|---|---|---|
| kglite_datasets-0.1.28-cp310-abi3-win_amd64.whl | CPython 3.10 | abi3 | Windows x86-64 | Details |
| kglite_datasets-0.1.28-cp310-abi3-manylinux_2_28_aarch64.whl | CPython 3.10 | abi3 | Linux glibc 2.28+ ARM64 | Details |
| kglite_datasets-0.1.28-cp310-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl | CPython 3.10 | abi3 | Linux glibc 2.17+ x86-64 | Details |
| kglite_datasets-0.1.28-cp310-abi3-macosx_11_0_arm64.whl | CPython 3.10 | abi3 | macOS 11.0+ ARM64 | Details |
| kglite_datasets-0.1.28-cp310-abi3-macosx_10_12_x86_64.whl | CPython 3.10 | abi3 | macOS 10.12+ x86-64 | Details |
Total release size: 11.7 MB
Release files / kglite_datasets-0.1.28.tar.gz
| Download URL | kglite_datasets-0.1.28.tar.gz |
|---|---|
| Size | 300.6 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
c18a75509cc83a29a4814bfb5f4d47f0cda1e5b8d879245238ce25cfbe5275f7
|
|
BLAKE2b-256 checksum How to use checksums |
3f8b930c292b693edeaff447dda200fd6e34b381de645928d778776af3c62935
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Oct 10, 2026.
Transparency logRelease files / kglite_datasets-0.1.28-cp310-abi3-win_amd64.whl
| Download URL | kglite_datasets-0.1.28-cp310-abi3-win_amd64.whl |
|---|---|
| Size | 2.3 MB |
| Tags | CPython 3.10 Windows x86-64 abi3 |
|
SHA-256 checksum How to use checksums |
54113d37ecbd3e80f2293a10f365e57efb37607b3a8a4cbc4100064de18a1f6a
|
|
BLAKE2b-256 checksum How to use checksums |
b1f39e1a7e1cfb82e0fde446e86ea88ce625545ea78a8a066bf02abb84388427
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Oct 10, 2026.
Transparency logRelease files / kglite_datasets-0.1.28-cp310-abi3-manylinux_2_28_aarch64.whl
| Download URL | kglite_datasets-0.1.28-cp310-abi3-manylinux_2_28_aarch64.whl |
|---|---|
| Size | 2.2 MB |
| Tags | CPython 3.10 Linux glibc 2.28+ ARM64 abi3 |
|
SHA-256 checksum How to use checksums |
ae2756b7a77ff24da1136b26d8eb7e690d7f747215263748ec38f9d65dc153cc
|
|
BLAKE2b-256 checksum How to use checksums |
4880980fab9eb1377bdae073503cb454cc92fe4d5be9900e8a1e2e49a64d8905
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Oct 10, 2026.
Transparency logRelease files / kglite_datasets-0.1.28-cp310-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
| Download URL | kglite_datasets-0.1.28-cp310-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl |
|---|---|
| Size | 2.4 MB |
| Tags | CPython 3.10 Linux glibc 2.17+ x86-64 abi3 |
|
SHA-256 checksum How to use checksums |
0f873b9953eaeb589a0213918ae60d811ff8bf50011ee818434801861517f66c
|
|
BLAKE2b-256 checksum How to use checksums |
5e08be402d9f82d6e8ae11bb05a884ee540bd827fa9324035b8a0912ac32c724
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Oct 10, 2026.
Transparency logRelease files / kglite_datasets-0.1.28-cp310-abi3-macosx_11_0_arm64.whl
| Download URL | kglite_datasets-0.1.28-cp310-abi3-macosx_11_0_arm64.whl |
|---|---|
| Size | 2.1 MB |
| Tags | CPython 3.10 abi3 macOS 11.0+ ARM64 |
|
SHA-256 checksum How to use checksums |
1abbe81c35ff1982affdb2b305cca2f7de5f36a2527322e862d85e9d399169a5
|
|
BLAKE2b-256 checksum How to use checksums |
03dedbe8af2923fd6892f0ec5f61dd002f658e924bb063ff034fc861d6578376
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Oct 10, 2026.
Transparency logRelease files / kglite_datasets-0.1.28-cp310-abi3-macosx_10_12_x86_64.whl
| Download URL | kglite_datasets-0.1.28-cp310-abi3-macosx_10_12_x86_64.whl |
|---|---|
| Size | 2.3 MB |
| Tags | CPython 3.10 abi3 macOS 10.12+ x86-64 |
|
SHA-256 checksum How to use checksums |
8daf5e193ec81ec002ecf2c6731358d1fa54f3014ccb5a2cfa3ea08376e49da2
|
|
BLAKE2b-256 checksum How to use checksums |
0dfe3f679acb539937c3c4ea88df04813cd0fbf57565eef1ffc5e1463dd48012
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Oct 10, 2026.
Transparency log
