A customer from Lithuania needed to replace their original solution based on the 360Works Scribe plug-in, which was used for automatically filling in data into Word and Excel templates and retrieving data back from these documents. The main issue was the slow processing speed for large documents and conflicts between Java version requirements for FileMaker Server, the Scribe plug-in, and other components.
We developed a standalone HTTP service written in Python that enables fast and reliable document generation and processing. Since the service runs as an HTTP server, FileMaker and other applications can easily use it via API through Insert from URL.
The service handles inserting formatted data into Word and Excel templates, as well as reading data from those documents. It can also generate PDF documents from Office files, insert text elements and images into PDFs at exact positions, encrypt and zip PDFs, and even create electronic documents compliant with Lithuanian government standards.
For converting documents to PDF, we use LibreOffice in command-line mode, which allows fast and precise PDF generation. Thanks to the use of technologies like Python and Docker, the service is easily scalable — you can run a single instance of DocGenerator for multiple FileMaker apps or multiple instances of DocGenerator for one more demanding FileMaker application. A major advantage is full platform independence (Windows, macOS, Linux) and no dependency on Java.
The entire solution resulted in reduced manual errors and a significant acceleration of processing — often saving several minutes to tens of minutes compared to the original solution.
