Open the files. Check the rows.

One fictional statement.
Every output to inspect.

Compare the source PDF with CSV, Excel and OFX files produced by Fulla’s actual exporters. No account or upload needed.

Download the complete sample pack

ZIP · fictional data · includes the method and validation results

What this example proves

Real export code. Authored sample rows.

Manually authored fictional rows passed to the current Fulla Ledger export functions. No extraction model or external accounting importer was run to create these downloads.

The PDF is clearly marked fictional. The files demonstrate format structure, monetary values and a review check. They are not a bank-issued statement, an extraction-accuracy benchmark, or evidence of a completed QuickBooks or Xero import. The existing product videos use their own fictional examples.

Individual files

Choose what you want to inspect.

  • PDF

    Fictional source statement

    One page with selectable text and eight fictional transactions.

    Download
  • JSON

    Expected approved rows

    Complete authored rows, metadata, stable IDs and printed original fields.

    Download
  • CSV

    Fulla CSV output

    Signed amounts, debit and credit columns, UTF-8 BOM and protected formula-leading text.

    Download
  • XLSX

    Fulla Excel output

    Transactions, Statement and Original Fields sheets with numeric monetary cells.

    Download
  • OFX

    Fulla OFX output

    OFX 1.0.2 SGML with stable FITIDs; routing-number placeholder 000000000.

    Download
  • JSON

    Worked sign correction

    The deliberately incorrect row, correction and before/after balance difference.

    Download
  • JSON

    Validation and file hashes

    Reconciliation, export checks, source fingerprints and SHA-256 file hashes.

    Download
  • TXT

    Method and limitations

    Reproduction commands, format details and limits of this fictional demonstration.

    Download

A worked review example

One reversed sign changes the check by $169.00.

The statement shows a $84.50 supplies debit. If a reviewer records it as a credit, the sign is wrong in both directions: the computed closing balance becomes $3,870.00.

Check the printed debit, correct the row to -$84.50, and recalculate. The rows now explain the printed closing balance of $3,701.00.

Deliberately introduced sign error for a worked review example, not an observed model error. Every downloadable export uses the corrected rows.

Source-supported correction

Illustrative wrong amount
+84.50
Reviewed amount
-84.50
Difference before
169.00
Difference after
0.00
Download the correction details

Carry the context with the rows

Know what each file contains.

Excel: three worksheets

The workbook has Transactions, Statement, Original Fields sheets. Monetary values are numeric cells. Compare the source-page references and printed original fields with the PDF.

Explore bank statement to Excel conversion

CSV: preserve the original

The full export includes review context. Make a separate mapped copy for an importer that expects fewer columns. Keep one date convention and check signed amounts, row counts and formula-leading text.

Prepare a CSV import

OFX: inspect the account

Open the text file to inspect currency, period, transaction amounts, stable FITIDs and closing balance. This fictional example uses a routing-number placeholder of 000000000. It is not a live-account import file.

See how PDF to OFX works

JSON: explicit expected rows

The authored reference preserves the expected row values and metadata. Compare it with a result you obtain separately; do not treat it as an observed model response.

Plan an API workflow

A reproducible format check.

The method file documents the generation commands. The validation record contains file hashes and reconciliation checks so you can confirm that the files belong to the same sample. No real customer data is included.

A zero balance difference is one check, not proof that a real statement was extracted completely. Compare every date, description, amount and source page before approving your own result.