def prepare_arf_calc_paths(corp, attrname, logstep=0.02): """ Calculates frequencies, ARFs and document frequencies for a specified corpus. Because this is quite computationally demanding the function is typically called in background by KonText. arguments: corp -- a corpus instance attrname -- name of a positional or structure's attribute logstep -- specifies how often (as a ratio of calculated data) should the logfile be updated """ outfilename = corplib.subcorp_base_file(corp, attrname).encode('utf-8') if os.path.isfile(outfilename + '.arf') and os.path.isfile(outfilename + '.docf'): return None elif hasattr(corp, 'spath'): return corp.spath else: return None