# written by Raynor Vliegendhart
# see LICENSE.txt for license information
import os
import sys
import codecs
from optparse import OptionParser
from Tribler.Core.CacheDB.sqlitecachedb import SQLiteCacheDB

def main():
    parser = OptionParser(usage='usage: %prog [OPTION]... SDB')
    parser.add_option('-o',
                      dest='output',
                      metavar='DIR',
                      default=os.getcwd(),
                      help='outputdir, default: cwd')
    
    options, args = parser.parse_args()
    
    if len(args)!=1:
        parser.print_help()
        exit()
    
    filename = args[0]
    outputdir = options.output
    
    if not os.path.exists(filename):
        print "*** Cannot find '%s'" % filename
        sys.exit(4)
    
    if not os.path.exists(outputdir):
        os.mkdir(outputdir)
    
    sqlite_db_path = sys.argv[1]
    sqlitedb = SQLiteCacheDB.getInstance(None)
    sqlitedb.initDB(sqlite_db_path)
    
    sql = """
        SELECT TF1.term || " " || TF2.term AS phrase,
               COUNT(*) AS freq
        FROM TorrentBiTermPhrase P, TermFrequency TF1, TermFrequency TF2
        WHERE P.term1_id = TF1.term_id AND P.term2_id = TF2.term_id
        GROUP BY term1_id, term2_id
        ORDER by freq DESC;
    """
    bigrams = sqlitedb.fetchall(sql)
    
    sql = """SELECT term, freq FROM TermFrequency WHERE freq > 1 ORDER by freq DESC;"""
    unigrams = sqlitedb.fetchall(sql)
    
    
    fh = codecs.open(os.path.join(outputdir, 'nbuzz_uni.txt'), 'w', 'utf8')
    for term, freq in unigrams:
        print >>fh, freq, term
    fh.close()
    
    fh = codecs.open(os.path.join(outputdir, 'nbuzz_bi.txt'), 'w', 'utf8')
    for phrase, freq in bigrams:
        print >>fh, freq, phrase
    fh.close()


if __name__ == '__main__':
    main()