Couchdb-python: How can I call functions in map_reduce function? -

i want call functions don't know how import package. tried follows failed. how can do? (i want use third-party package analyzing each doc, code below test)

please, if know answer...

import couchdb db import datetime  couch = db.server()  d1 = couch['test']  def map(doc):     text = doc['text']     ti = doc['timestamp_ms']     ti = ti[:10] + '.' + ti[10:]     datearray = datetime.datetime.utcfromtimestamp(float(time))     if (datearray.time().hour < 12):             yield ["am"], text     else:             yield ["pm"], text  row in d1.query(map, descending=true, language='python'):         print row.key, row.value 

you can use datetime inside map function if import datetime inside map function. example -

     def map(doc):         text = doc['text']         ti = doc['timestamp_ms']         ti = ti[:10] + '.' + ti[10:]         from datetime import datetime         datearray = datetime.utcfromtimestamp(float(time))         if (datearray.time().hour 

however, cannot load third-party libraries in map function. tried importing nose test library , saw in couch logs -

     [[{initial_call,        {couch_os_process,init,['argument__1']}},       {pid,<0.14643.4>},       {registered_name,[]},       {error_info,        {exit,         {function_clause,          [{couch_os_process,handle_info,            [{#port<0.13927>,              {data,               {eol,                <<"{\"log\": \"traceback (most recent call last):\n  file \\"/usr/local/lib/python2.7/dist-packages/couchdb/view.py\\", line 79, in map_doc\n    results.append([[key, value] key, value in function(doc)])\n  file \\"\\", line 5, in map\nimporterror: no module named nose\n\"}">>}}},             {os_proc,"couchpy",#port<0.13927>,              #fun,              #fun,5000}]},           {gen_server,handle_msg,5},           {proc_lib,init_p_do_apply,3}]},         [{gen_server,terminate,6},          {proc_lib,init_p_do_apply,3}]}},
  

you might able if install third party library directly on host python installation, not in virtual environment. however, not recommend practice. use virtual env instead always.