ScanPdf 2.0.12: import of sales Orders and automatic OCR for unreadable PDFs

What this release brings

ScanPdf imports the PDFs received by a company (invoices and purchase orders, delivery notes, bank statements, payslips) and creates the corresponding Dolibarr objects. Version 2.0.12 adds a new document type and makes scanned PDF reading more reliable.

Sales Orders

The PDF order form sent by a customer becomes a Dolibarr sales Order:

• the customer is recognized in the PDF, and created on import if unknown;
• a preview shows, before creation, the customer, the items, the quantities, the tax-exclusive prices, and the total, along with the origin of each price;
• a warning flags a line where the price differs from your sales price list;
• the PDF can provide a single order or one order per line.

An external module can also hook into the import workflow to choose the item and price for a line from its own repository, or reject an inconsistent import.

Scanned PDFs: automatic OCR

Some PDFs display correctly but only contain unreadable text. ScanPdf now measures the quality of the text layer and sends these documents for character recognition instead of producing an empty extraction.

Security and reliability

• Documents, extraction templates, and learned patterns partitioned by entity on multi-entity Dolibarr.
• Dolibarr permissions checked before creating a third party, an invoice, or an order.
• OCR password encrypted, remote services called over https only, secrets removed from logs.
• Automatic publication of templates on the hub disabled by default.
• Brazilian Portuguese translation, translations completed in all languages.

Details of new features and fixes are on the Version 2.0.12 release notes page.

A demo set

The new Demo tab installs fictitious vendors and a customer with their sample PDFs: import one and see ScanPdf recognize the issuer and fill in the fields.

Try it

Try ScanPdf on a Dolibarr instance set up in under 20 seconds: dolitest.fr/scanpdf