To help isolate the root cause, can you create a new TestComplete project solely to check out the `Preparing an object recognition hint.` step that takes long time?
Do you get the same slow behavior in the new project?
Also if you attempt to map a new object in the NameMapping, does it take a long time?
Adding objects with ObjectSpy all goes fine. But the tree is massive. the NameMapping.tcNM file is 8000 lines. and the TypeInfo.dat is 2.6MB
Your NameMapping is large alright but there are too many variables to consider such as structure and extended find. As scot1967 suggested and as the following post Name Mapping too big ? | SmartBear Community pointing out, you can try it out in a new project.
When it happens, try to reproduce in new project by importing the NameMapping and isolating and copying the code to simulate. Once reproduced, experiment with the extended find and cleaning up the NameMapping.
Hope this helps.
When we ran into this issue. The solution was to limit search depth in our tests. This required specifying parent objects no more than a 8 or 10 nodes away from the target. This also helped to make object identification more stable as sometimes objects may have the same properties and TC will do a first match so your identification parameters have to be very specific.
