Twitterで声優のフォローリストを取得する

こちらを参考に一回200人の取得をAPIに引っかからないように取得する。

フォローしている人(friends)もしくはフォローされている人(followers)で別個に取得できる。有名人ならたいていfollowersが万人規模になるので、取得するだけでくっそ時間がかかる。

# 待ち時間をプログレスバーで
import sys, time
rng_time = 60*15
wait_time = 1
def LimitError(rng_time = 60*15, wait_time = 1):
	for i in range(1, rng_time + 1):
		rate = i * 71 / rng_time
		meter = "\r|%s%s| %d秒" % ("=" * rate, ' ' * (71 - rate), (rng_time - i))
		sys.stdout.write(meter)
		sys.stdout.flush()
		time.sleep(wait_time)
	
	sys.stderr.write('\n')

sys.stderr.write('\n')

fnum = 200 # 一回の取得数
query_per_sec = 65 #15*60/180.0 # API 待機時間
func = "friends" # friends か followers か
wd_out = "/" + func + "/"

# cv というのがアカウント名のリスト
for user in cv:
	try:
		suser = twitter.show_user(screen_name=user)
		e = ["0"]
		LimitError(rng_time = query_per_sec, wait_time = 1)
	except TwythonError, e:
		print(user)
	except TwythonRateLimitError, e:
		print("TwythonRateLimitError")
		LimitError(rng_time = 60*15, wait_time = 1)
	
	if list(e)[0].find("4") < 0:
		pnum = int(math.ceil(float(suser[func + "_count"]) / fnum)) # 人数
		print("Processing " + user + " (" + str(cv.index(user)+1) + "/" + str(len(cv)) + ") " + str(pnum))
		pages = []
		oldpages = []
		for i in range(pnum):
			pages.append("p"+str(i+1))
			oldpages.append("p"+str(i))
		
		if pnum == 0: # 0 人なら書きだすだけで終了する。
			w0 = open(wd_out + user + ".txt", "w")
			w0.close()
		
		p0 = { "next_cursor": -1 } # So the following exec() call doesn't fail.
		# friends もしくは followers の取得
		for i in range(pnum):
			try:
				exec(pages[i]+" = twitter.get_" + func + "_list(screen_name=user, count=fnum, skip_status=1, cursor="+oldpages[i]+"['next_cursor'])")
				LimitError(rng_time = query_per_sec, wait_time = 1)
			except TwythonRateLimitError, e:
				print("TwythonRateLimitError")
				LimitError(rng_time = 60*15, wait_time = 1)
				pass
			except TwythonAuthError, e:
				print("TwythonAuthError")
				LimitError(rng_time = query_per_sec, wait_time = 1)
		
		if i == pnum-1:
			res = []
			for p in range(pnum):
					try:
							exec("for i in range(fnum): res.append("+pages[p]+"['users'][i])")
					except(IndexError):
							pass
			
			a = map(lambda x: x["screen_name"], res)
			w0 = open(wd_out + user + ".txt", "w")
			w0.write("\t".join(a) + "\n")
			w0.close()