Apache PDFBox - merge documents without loading objects to memory

Question

I'm trying to merge PDF documents using PDFMergerUtility. I was thinking about mergeDocuments() method but when I looked into the source code, the files are loaded as in-memory PDDocuments, which I'm trying to avoid, because my application works with huge images that are exported to pdf and creating PDDocument containing these images causes OutOfMemoryError: Java heap space.

Is there any way to merge these documents without loading the whole objects to memory, using PDFBox?

Answer 1

The method of PDFMergerUtility that allows for merging documents, accepts a parameter that allows you to define what strategy may be used to manager memory during the merge.

There are 3 strategies: using the main memory, using temporary files and a mix of the other 2; additionally, you can tune up to what limit one strategy or the other may be used.

In your case, you should go for temporary files only - this can be done as follows [assuming the instance of your PDFMergerUtility object is called pdfMerger] :

pdfMerger.mergeDocuments(MemoryUsageSettings.setupTempFileOnly().streamcache)

(the version of Pdfbox used in this solution is 3.0.5)