Most time consuming parts of the code.
NOTE:
  * 1st line is irrelevant
  * one execfile is used by cProfile, execfile of the ontology takes 6ms
  * importing zeitgeist.datamodel takes 22 ms (line 3)
  * there seem to be two methods which take the most time:
    - Symbol.__getitem__()
    - Symbol._resolve_children()

markus@thekorn ~/devel/zeitgeist/ontology_definition % /usr/lib/python2.6/cProfile.py --sort cumulative zeitgeist/datamodel.py
         18368 function calls (13136 primitive calls) in 0.030 CPU seconds

   Ordered by: cumulative time

   ncalls  tottime  percall  cumtime  percall filename:lineno(function)
        1    0.000    0.000    0.030    0.030 <string>:1(<module>)
      2/1    0.011    0.006    0.030    0.030 {execfile}
        1    0.001    0.001    0.022    0.022 datamodel.py:23(<module>)
      106    0.001    0.000    0.016    0.000 datamodel.py:157(_resolve_children)
      188    0.001    0.000    0.014    0.000 datamodel.py:197(__getitem__)
      423    0.000    0.000    0.013    0.000 {next}
        1    0.000    0.000    0.012    0.012 zeitgeist.py:4(<module>)
       80    0.000    0.000    0.012    0.000 datamodel.py:147(__init__)
      204    0.002    0.000    0.007    0.000 datamodel.py:209(<genexpr>)
      219    0.005    0.000    0.006    0.000 datamodel.py:203(<genexpr>)
7655/2424    0.004    0.000    0.004    0.000 datamodel.py:269(iter_all_children)
     3982    0.001    0.000    0.001    0.000 datamodel.py:230(uri)
      121    0.000    0.000    0.001    0.000 {filter}