Shell Programming and Scripting

1999001.txt
1999002.txt
1999003.txt
1999004.txt ...
1999365.txt

desired output:
19990101.txt
19990102.txt
19990103.txt
19990104.txt ...
19991231.txt

1999001.txt renamed to 19990101.txt
12.3 124.4 2.0
12.1 124.2 1.2
12.4 123.5 1.5

1999002.txt renamed to 19990102.txt
12.3 124.4 1.8
12.1 124.2 2.2
12.4 123.5 1.2

1999003.txt renamed to 19990103.txt ...
12.3 124.4 1.5
 12.1 124.2 2.4
 12.4 123.5 1.5

1999007.txt renamed to 19990107.txt
12.3 124.4 1.2
12.1 124.2 2.6
12.4 123.5 1.8

hence, in Jan and Feb for instance, I would have resulting averaged files like the ones below:

199901_wk01.txt average of files in Jan 1999 from day 1-7
199901_wk02.txt average of files in Jan 1999 from day 8-14
199901_wk03.txt average of files in Jan 1999 from day 15-21
199901_wk04.txt average of files in Jan 1999 from day 22-28
199901_wk05.txt average of files in Jan 1999 from day 29-31

199902_wk01.txt average of files in Feb 1999 from day 1-7
and so on

#! /usr/bin/python

import datetime
import glob
import os


# $ cal 10 1999
#
#    October 1999
#Su Mo Tu We Th Fr Sa
#                1  2
# 3  4  5  6  7  8  9
#10 11 12 13 14 15 16
#17 18 19 20 21 22 23
#24 25 26 27 28 29 30
#31
#
# 1999-10-28 return 5
def getwk(y,m,d):
        blanks=datetime.date(y,m,1).weekday()
        return (d+blanks)/7+1


average=dict()
for source in glob.glob('1999[0-3][0-9][0-9].txt'):
        year=int(source[0:4])
        nday=int(source[4:7].lstrip('0'))
        dt=datetime.timedelta(days=nday-1)
        d=datetime.date(year,1,1)+dt
        target=str(d)+'.txt'

        os.rename(source,target)

        month=int(target[5:7].lstrip('0'))
        day=int(target[8:10].lstrip('0'))
        week=getwk(year,month,day)

        # 199903_wk03.txt
        target_wk=target[0:4]+target[5:7]+'_wk'+'%02d'%week+'.txt'

        for line in open(target):
                try:
                        average[target_wk].append(float(line.split()[2]))
                except:
                        average[target_wk]=[float(line.split()[2])]

# write out the average to week files.
for i in average:
        avg=float(sum(average[i]))/len(average[i])
        f=open(i,'w')
        f.write('%.2f\n'%avg)
        f.close()

#! /bin/bash

rm -f 1999*.txt
seq -w 1 365 | while read d
do
        echo "12.3 124.4 $RANDOM
12.1 124.2 $RANDOM
12.4 123.5 $RANDOM" > 1999$d.txt
done