Read a source-contract demo independently

Synthetic Stage 0 demonstration; no live OSDU or application connection required. These are authored test values, not NLOG observations or production well measurements.

1. Download the synthetic demo ZIP.
2. Extract it into a new folder and open a terminal there.
3. Run the included independent reader:

python3 verify.py .

Only Python’s standard library is needed. No Ophiolite account, package, token or license is required. The script checks payload hashes and reads the CSV/JSON data. It prints PASS when depth, values, missing data and SQL NULL/empty text match.

File	What to inspect
README.txt	Plain-language outcomes and limitations
synthetic.las	Original gamma values 10, missing, 30 at depths 100–102 m
edited.las	First gamma sample changed to 12.25; missing sample retained
values.csv	Values parsed from the edited LAS; readable in a spreadsheet
synthetic-wells.sqlite	Two synthetic well rows, including duplicate display names with different IDs
sql-source.json	Original rows, mapped values and declared scientific context
report.json	Mapping/fidelity evidence and illustrative policy decisions

Gamma values retain the declared gAPI label. The synthetic reference failure case shows inspection allowed and a dependent calculation refused. It does not assert that a particular customer’s reference catalogue is available or unavailable. Depth reference remains unknown; declared units do not establish a vertical datum.

The unchanged export test preserves original bytes. The edited export reports formatting loss; exact source formatting is not promised after rewriting. A source read grant alone fails the retained-sharing policy demonstration. Its success case uses explicitly supplied synthetic grants, not a running sharing service.

The packaged reader verifies the listed file hashes, not a cryptographic signature of the publisher. A checksum is neither permission nor proof of scientific correctness. If verification fails, extract a fresh copy before retrying; do not edit expected hashes.

Developers with the experimental Connectors environment can regenerate the data:

python -m asset_connectors.contract_demo /tmp/new-contract-demo

Choose a directory that does not exist. This generator uses the existing LAS/SQL readers and never connects to customer data. See source qualification.

For an explicit zero-versus-missing example, use the manual source setup fixtures. A blank numeric CSV field means missing in these examples; a written 0 means zero. Do not let another reader silently fill blanks. The original LAS retains its declared NULL marker; neither the table preview nor the curve renderer interpolates gaps.