We have identified another threading problem in the latest release (10.0.0). When extracting text from PDF documents using multiple threads, the CPU will spike to 100% randomly and never finish. In most cases this happens when multiple threads are extracting text from documents with many pages. Please note that version 9.7.0 works fine in multi-threaded environments.
The problem appears to happen either when adding the TextAbsorber to the Page object, or when retrieving the text from the TextAbsorber.Text property.
I have added a sample application that will demonstrate the problem. I have also include many PDFs in the working folder. 9 times out of 10, CPU usage will go to 100% (and stay there) before finishing the test when not using a lock. If a lock is used, the test will run fine (but very slowly). The code is commented with more information.
Please note, you will need to supply a valid license. It will not exhibit the problem with the trial version.
Thank you,
Scott