Withdrawn standalone offer · v1.0.0 · No new orders

Catch silent table changes before the next run.

The implemented tool saves a baseline for one local CSV or JSONL file, then classifies the next copy as identical, the same retained shape, or drift—without copying row values into the report.

Inspect withdrawal record Inspect synthetic evaluation Browse current offers

No new standalone orders are accepted. The private implementation, wheel, tests, and delivery archive remain unpublished and are not relicensed by this withdrawal. The public evaluation contains synthetic records only.

01 / DecisionStandalone sale withdrawn

The workflow was not sufficiently differentiated from capable free validation, hashing, and AI-assisted scripting substitutes under a manual purchase and delayed-delivery model.

02 / What remainsEvidence, not executable code

The strict schemas, synthetic inputs, expected receipt, exact/shape/drift comparisons, manifest, and checksums remain public for inspection.

03 / What did not happenNo free-code release

The private implementation remains unpublished. Withdrawal grants no software license and makes no customer, defect, refund, security, or demand claim.

What it answers

Is this the same file, the same recorded structure, or a changed table?

Use it before a handoff, import, recurring analysis, model-input refresh, or automation step when a data-quality platform is excessive but a visual spot-check is too weak. The result names the changed dimensions without printing row values.

Former technical fit

The intended use fit four narrow conditions.

01

The source is one local table

You have one bounded UTF-8 CSV or JSON Lines file—not a database, workbook, Parquet dataset, or remote URL.

02

Structural evidence is enough

Counts, ordered names, type categories, missing/null/blank counts, duplicate counts, and timestamp grids answer the immediate question.

03

Values should stay out

You do not want rows, samples, extrema, quantiles, frequent values, or individual row hashes written into the receipt.

04

You can declare the parsing boundary

File suffix, CSV delimiter and null markers, plus resource ceilings can be explicit and stable.

Exact / shape / drift workflowInspect parsing rules, comparison exits, release files, redaction tests, and support.

Former licensed runtime

Create once. Verify or compare later.

python -m tabular_shape_receipt create input.csv --out receipt-pack --label vendor-feed
python -m tabular_shape_receipt verify receipt-pack
python -m tabular_shape_receipt compare receipt-pack current.csv

compare exits 0 for EXACT_MATCH or SHAPE_MATCH, 2 for DRIFT, and 1 for invalid input, an invalid pack, or an operational error.

Former delivery contents

What the withdrawn package contained.

  • Source + wheelStandard-library package under an individual or one-organization commercial license.
  • Three output schemasDraft 2020-12 contracts for receipts, verification results, and comparisons.
  • Twenty-two passing testsFormat, limit, determinism, redaction, tamper, interruption, comparison, schema, and CLI coverage; one host-conditional symlink case.
  • Synthetic evaluationInspectable exact, shape-match, and drift cases with no executable implementation.
  • Checksums + supportExact release identities and the former bounded installation-support term.
Public evaluation recordsOpen the exact evaluation hash, synthetic receipt, public schemas, and checksum record.
Evaluation ZIP SHA-256CF7EE9AC62B9FC983B1E89EE1EF3E07EE8E135FFDB3706AE7A3DE33F73C79546

Commercial correction

Not sold separately.

No new orders are accepted. Use a current validator or a reviewed local script when either answers the job; inspect the current catalog for separately differentiated KaldenFrame offers.

Browse current products Read withdrawal JSON Read retained terms record