I am using hunspell 1.2.1 to check wrong words in a file from a software.
I am issuing the command
cmd = "hunspell -i utf8 -d " + mydictionary

and sending the command into terminal through

Popen(cmd, stdin=PIPE, stderr=STDOUT, stdout=PIPE, shell=isinstance(cmd, basestring)).communicate(message)[0]

here message is the message that i want to check the spelling.

This command causes loading the whole dictionary each time i want to check a message in the file and thus makes my spell checking operation very slow.
Is there any other way of using hunspell to make spell checking operation more faster. it will be better if i load the dictionary for only one time and allow checking the spelling against the loaded dictionary rather than loading the whole dictionary each time.