Card numbers: fake and obviously not real (0000, 1111, 2222).
No sensitive identifiers: no addresses, emails, phone numbers, account numbers.
Files
extracted.csv
Semicolon-separated (;) to match the CLI output.
Header must match the extracted CSV schema exactly.
Do not hardcode the full header in this README; it is enforced by Go tests.
mappings.v1.json
Minimal exact-match mappings used by the dashboard.
What this dataset should cover (intentionally, but not rigidly)
Multiple statement months (at least 2 different CloseDate months).
Installments (at least one CardMovement with CurrentInstallment/TotalInstallments set).
Refund pairs:
one matching positive/negative pair across different months, and
one matching positive/negative pair within the same month.
Mapping gaps (to feed DQ):
at least one unmapped category, and
at least one unmapped owner.
Note: Parse-invalid fixtures (bad date formats, invalid MovementType strings, etc.) should live as additional files in this folder (e.g. extracted.invalid_*.csv) so the main demo dataset stays usable for the hosted demo.
Planned simple additions (keep each addition small and purposeful)
USD examples (at least one CardMovement with AmountUSD != 0).
More banks/card companies (additional CardCompany / Bank combinations).
var (
// ExtractedCSV is a synthetic/anonymized extracted CSV fixture.////go:embed extracted.csv
ExtractedCSV string// MappingsV1JSON is a minimal mappings fixture consumed by the dashboard.////go:embed mappings.v1.json
MappingsV1JSON string
)