- added graceful shutdown (with one ^C) and forced shutdown (two ^C)
- added optional loglevel to IgnoreRequest exception
- calling Spider Manager close_domain() function from engine to avoid race
  conditions caused by signal dispatch order
- moved twisted reactor controlling logic outside the engine and into the
  scrapy manager
- made log.err function respect the current log level filter

fix bug in scrapy shell which was hiding the objects fetch/view/shelp when started without a url as argument

--HG--
extra : rebase_source : 976f3ca2fccf185399a733904b9763498ebf43f6

--HG--
extra : rebase_source : cd0d5f560dcde0003974e360282a233ba387c5be

--HG--
extra : rebase_source : cd450c66be44381e905264ecff3f709bb9d302d4

--HG--
extra : rebase_source : 71228e29b0a969df06ac62d90d6e3cefc03ecdfc

--HG--
rename : scrapy/stats/corestats.py => scrapy/contrib/corestats.py

--HG--
rename : scrapy/contrib/spider/profiler.py => scrapy/contrib_exp/spiderprofiler.py

--HG--
rename : docs/experimental/exporters.rst => docs/topics/exporters.rst

- calling scrapymanager.configure() for all commands
- finally added some unittests to check cmdline behaviour!

moved engine.getstatus() method to scrapy.utils.engine function, to leave reporting logic out of engine code. added est() shortcut to telnet console

MemoryUsage: changed .virtual property to methodd. SpiderProfiler: removed dependency on MemoryUsage extension