Parcel Billable Weight Evidence Report

Compare a batch of supplied parcel measurements and invoice billed weights under your own explicit calculation policy. The report separates dimensional arithmetic, actual weight, changes between declared and audited measurements, and invoice comparison differences. It provides source references for review without carrier account credentials.

Use it after a packing measurement review or when a batch of weight adjustments arrives. Upload normalized arrays or mapped CSV exports, declare the applicable units and rounding rules, then inspect the complete report and download evidence for your shipping team. A later batch is a new report.

What the report calculates

For each resolved measurement basis, the Actor rounds each dimension according to your policy, multiplies the rounded length, width and height, and divides that volume by your dimensional divisor. It rounds both actual and dimensional weight up to your billable weight increment, then takes the greater result. Arithmetic uses exact decimal and rational values; a nonterminating dimensional result appears as a fraction such as 768/139, rather than a rounded approximation. Final billable weights are exact decimal strings.

Invoice weight is compared separately against declared and audited calculations. The signed difference is invoice weight minus computed weight. A difference equal to the declared tolerance is within tolerance. The report also shows whether supplied dimensions or actual weights changed, even when both measurements yield the same billable weight.

The included synthetic example has four packages: an actual-weight-dominant package with changed measurements and invoice differences, a matching actual-weight-dominant package, a matching dimensional-weight-dominant package, and a package with missing required evidence. The example policy is illustrative and is not an authoritative carrier billing rule.

Input

Provide snapshotId, requiredBases, evidenceCoverage, csvDelimiter, and a policies array. Each policy needs a unique scopeId, policyRef, dimensionUnit (in or cm), weightUnit (lb or kg), dimensionalDivisor, dimensionRounding (none, ceil, or nearest-half-up), dimensionIncrement, billableWeightIncrement, and tolerance. The divisor must have the units of your chosen dimension unit cubed per chosen weight unit. There is no implicit divisor or carrier selection. Dimension rounding none preserves the original measurements and ignores the supplied positive dimension increment. Nearest-half-up rounds exact halfway values upward; it may quantize a very small positive dimension to zero. These are arithmetic policies, not validations of physical package size.

Measurement rows require recordId, packageId, scopeId, basis (declared or audited), length, width, height, dimensionUnit, actualWeight, weightUnit, recordedAt, and sourceReference. Invoice rows require recordId, packageId, scopeId, billedWeight, weightUnit, and sourceReference. Timestamps must be real UTC timestamps in YYYY-MM-DDTHH:mm:ss[.sss]Z form. Identifiers are literal strings, preserving leading zeroes; omit customer names and addresses.

All arithmetic input values are decimal strings with a dot, at most six fractional places, and a maximum of 999999.999999. Dimensions, divisors and increments must be positive. Weights and tolerance may be zero. No grouped numbers, signs, exponents or unit conversion are accepted. Each measurement and invoice unit must match its declared scope policy.

Supply measurements and invoices arrays, including explicit empty arrays when appropriate, or their corresponding measurementsCsv and invoicesCsv alternatives. Remove an array when supplying its CSV. Optional measurementsColumns and invoicesColumns map every canonical field to a distinct CSV header. UTF-8 BOM, quoted fields, and comma, semicolon or tab delimiters are supported. Duplicate or unknown headers, irregular rows and ambiguous numbers are rejected. Inputs are bounded to 4 MB and 5,000 combined policy, measurement and invoice records. At least one evidence row is required. Larger or unusually wide outputs may exceed the report limits; split those batches.

Evidence, uncertainty and downloads

Repeated package/basis measurements remain unresolved, including identical repeated rows. Duplicate invoice rows are not summed or selected. Repeated record IDs or policy scopes are rejected. A package appearing in different scopes cannot be joined across those scopes. Missing policies, invoice weights or caller-required measurement bases remain unknown; missing optional bases are marked not provided. Incomplete or unknown batch coverage prevents an overall consistent conclusion.

One dataset item contains the complete report, package evidence rows, policy snapshot and traceable findings. Download full JSON, flat parcel CSV, findings CSV, and standalone HTML. CSV cells neutralize spreadsheet formula prefixes; HTML escapes supplied text and uses no external resources. Source references identify the original array position or CSV record and retain your own evidence reference. Cloud input and results are stored by Apify; use only exports you are authorized to process.

Price and limits

The charge is $0.25 per completed report, with platform usage included. A report is generated and all export size limits are checked before its event is charged. Invalid inputs, insufficient event budget, or oversized reports produce no completed-report event. The full JSON must be under 8 MB and each export under 9 MB. A resumed run reuses its saved report and repairs partial exports without another dataset item or report charge, after checking input and saved-result integrity. A new run is a new billable report.

This Actor checks supplied arithmetic evidence. It does not validate physical measurements, confirm a carrier contract, calculate surcharges, zones, freight classes, shipping prices or refund entitlement, or identify fraud. A higher invoice weight may be explained by excluded rules or missing evidence. No PDF/OCR, carrier login, dispute submission, shipping purchase or account changes are performed. Human review is required before acting on findings.