파이썬으로 월별 통계 내기...

좋은 아이디어를 가지고 월별통계를 볼 수 있는 프로그램으로 파이썬과 wget으로 만들어보았습니다.
wget으로 만든 이유는 이런걸 하는 전용 프로그램이 있는데 굳이 파이썬의 urllib같은걸로 삽질할 필요가 없을 것 같아서 입니다.
Julian Date 모듈은 웹에서 찾아보니 쓸만한게 있더군요.(y,m,d형태로 형식만 바꿨습니다.)

또한 UTF-8환경에서 구동시키려면 iconv를 사용하여 이래저래 컨버전이 필요하더군요.

=======================================================
#!/usr/bin/python
import julian

def date_process(process_dt):
year,month,day=julian.DateToYMD(process_dt)
month+=1
if (month>12):
year+=1
month=1
gm_date=(year,month,day,0,0,0,0,0,0)
return "20" + time.strftime("%y%m%d",gm_date)

if __name__ == '__main__':
import os,sys,re,time
if len(sys.argv) < 4:
print 'Usage: ' + sys.argv[0] + 'from_dt(20050201) to_dt(20050301) keyword'
sys.exit()
start_time = time.mktime(time.strptime(sys.argv[1][2:], "%y%m%d"))+24*60*60/2
end_time = time.mktime(time.strptime(sys.argv[2][2:], "%y%m%d"))+24*60*60/2
keyword = sys.argv[3]
print keyword + " 에 대한 " + sys.argv[1] + "~" + sys.argv[2] + "까지의 검색인덱스 월별 건수"
while(start_time <= end_time) :
query_dt = "20" + time.strftime("%y%m%d",time.gmtime(start_time))
from_dt=julian.Julian(query_dt)
to_dt_time=date_process(query_dt)
to_dt=julian.Julian(to_dt_time)
query = "'http://www.google.com/search?hl=ko&num=10&q=\"" +keyword+'\" daterange:'+`from_dt`+'-'+`to_dt` + "'"
print query
ret_val = os.popen("echo " + query + " | iconv -f utf-8 -t cp949 | wget -i - -U 'Mozilla' -q -O - | iconv -f cp949 -t utf-8").read()
p=re.compile("<b>[0-9].*</b>개")
try:
count=p.search(ret_val).group()[3:][:-7] print `julian.JulianToYearMonthDay(from_dt)` + "~" + `julian.JulianToYearMonthDay(to_dt)` + "=" + count
except AttributeError:
print "Error or 0"
start_time = time.mktime(time.strptime(to_dt_time[2:], "%y%m%d")) + 24*60*60/2
==================================================================

julian.py
==================================================================
def NumDaysInMonth(year, month):
if month == 2:
if IsLeapYear(year):
return 29
else:
return 28
elif month == 9 or month == 4 or month == 6 or month == 11:
return 30
elif month == 1 or month == 3 or month == 5 or month == 7 or \
month == 8 or month == 10 or month == 12:
return 31
else:
raise "Bad month"

def JulianToYearMonthDay(julian_day):
if julian_day < 0: raise "Bad input value"
jd = julian_day + 0.5
Z = int(jd)
F = jd - Z
A = Z
if Z >= 2299161:
alpha = int((Z - 1867216.26)/36254.25)
A = Z + 1 + alpha - int(alpha/4)
B = A + 1524
C = int((B - 122.1)/365.25)
D = int(365.25 * C)
E = int((B - D)/30.6001)
day = B - D - int(30.6001 * E) + F
if E < 13.5:
month = int(E - 1)
else:
month = int(E - 13)
else:
month = int(E - 13)
if month > 2.5:
year = int(C - 4716)
else:
year = int(C - 4715)
return year, month, int(day)

def DayOfYear(year, month, day):
if IsLeapYear(year):
n = int((275*month)/9 - ((month + 9)/12) + int(day) - 30)
else:
n = int((275*month)/9 - 2*((month + 9)/12) + int(day) - 30)
if n < 1 or n > 366: raise "Internal error"
return n

def DayOfWeek(year, month, day):
julian = int(JulianAstro(month, int(day), year) + 1.5)
return julian % 7

def IsLeapYear(year):
if (year % 400 == 0) or (year % 4 == 0 and year % 100 != 0):
return 1
else:
return 0

def IsValidDate(year, month, day):
'''Returns true if the year is later than 1752 and the month and day
numbers are valid.
'''
if month < 1 or month > 12: return 0
'''
if month < 1 or month > 12: return 0
if int(month) != month : return 0
if year < 1753 : return 0
if day < 1.0 : return 0
if int(day) != day:
if month == 2:
if IsLeapYear(year):
if day >= 30.0: return 0
else:
if day >= 29.0: return 0
elif month == 9 or month == 4 or month == 6 or month == 11:
if day >= 31.0 : return 0
else:
if day >= 32.0 : return 0
else:
if month == 2:
if IsLeapYear(year):
if day >= 29 : return 0
else:
if day >= 28 : return 0
elif month == 9 or month == 4 or month == 6 or month == 11:
if day >= 30 : return 0
else:
if day >= 31 : return 0
return 1

def JulianAstro(year, month, day):
"Note that day can be either an integer or a float."
if month < 3:
year = year - 1
month = month + 12
year = year - 1
month = month + 12
julian = int(365.25*year) + int(30.6001*(month+1)) + day + 1720994.5
tmp = year + month / 100.0 + day / 10000.0
if tmp >= 1582.1015:
A = year / 100
B = 2 - A + A/4
julian = julian + B
return julian * 1.0

def Julian(year, month, day):
return int(JulianAstro(year, month, day) + 0.55)

def DateToYMD(date):
import string
year=string.atoi(date[0:4])
month=string.atoi(date[4:6])
day=string.atoi(date[6:8])
return year,month,day

def Julian(date):
year,month,day=DateToYMD(date)
return int(JulianAstro(year,month,day) + 0.55)

if __name__ == "__main__":
Test()