I have discovered a state problem that I did not understand, until I tried to put images in my document with the text.
It appears that you have to create all of the PDJpeg objects after the PDDocument is created and BEFORE you try to create any text using the PDPageContentStream object. When I tried to create the PDJpeg object in the same part of my code where I was creating text, the text PDF stream was getting mixed in with the "stream" where the XOjbect is defined. This situation makes some of the text disappear, and the PDF reader throws errors trying to process the image. This issue begs the question of how you create mixed image/text documents "on the fly", where you don't know whether you will embed an image, or where you don't know the name of the image file until the instant you have to create the text stream. I can provide code, if desired, but I used the ImageToPDF example as a base for trying to also write some text along with the image created in this code.
