##################################
#                                #
# Last modified 2026/02/18       # 
#                                #
# Georgi Marinov                 #
#                                # 
##################################

import sys
import string
import math
import os
from sets import Set

def run():

    if len(sys.argv) < 4:
        print 'usage: python %s  datafilename labelIDs valueIDs outfilename' % sys.argv[0]
        print 'format of IDs: either comma separated or start-end (including end)'
        sys.exit(1)

    datafilename = sys.argv[1]
    LabelIDfields=[]
    if '-' in sys.argv[2]:
        fields1=int(sys.argv[2].split('-')[0])
        fields2=int(sys.argv[2].split('-')[1])
        for ID in range(fields1,fields2+1):
            LabelIDfields.append(ID)
    else:
        fields = sys.argv[2].split(',')
        for ID in fields:
            LabelIDfields.append(int(ID))
    ValueIDfields=[]
    if '-' in sys.argv[3]:
        fields1=int(sys.argv[3].split('-')[0])
        fields2=int(sys.argv[3].split('-')[1])
        for ID in range(fields1,fields2+1):
            ValueIDfields.append(ID)
    else:
        fields = sys.argv[3].split(',')
        for ID in fields:
            ValueIDfields.append(int(ID))
    outfilename = sys.argv[4]

    outfile = open(outfilename, 'w')

    if datafilename.endswith('.bz2'):
        cmd = 'bzip2 -cd ' + datafilename
    elif datafilename.endswith('.gz') or datafilename.endswith('.bgz'):
        cmd = 'gunzip -c ' + datafilename
    elif datafilename.endswith('.zip'):
        cmd = 'unzip -p ' + datafilename
    else:
        cmd = 'cat ' + datafilename
    p = os.popen(cmd, "r")
    line = 'line'
    while line != '':
        line = p.readline().strip()
        if line == '':
            break
        if line[0]=='#':
            continue
        fields = line.strip().split('\t')
        outline=''
        for ID in LabelIDfields:
            outline=outline+fields[ID]+'\t'
        score=0
        for ID in ValueIDfields:
            score += float(fields[ID])
        outline=outline+str(score)
        if outline.endswith('.0'):
            outline=outline[0:-2]
        outfile.write(outline+'\n')

    outfile.close()
        
run()

