You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
pixiv/PixivSearch/dao/Comment.py

120 lines
4.0 KiB

6 years ago
import json
import os
import shutil
import threading
from concurrent import futures
import requests
from lxml import etree
class Comment:
lock = threading.Lock() # 多线程全局资源锁
def __init__(self, keywords_=None) -> None:
super().__init__()
self.obj = {'data': {}, 'flag': False}
self.keywords = keywords_
# 获取番剧合集弹幕排行榜
def getCommentSort(self, cids):
urls = []
for cid in cids:
urls.extend(getCidUrls(cid))
with futures.ThreadPoolExecutor(32) as executor:
executor.map(self.count, urls)
for index, data in enumerate(
sorted(self.obj["data"].items(), key=lambda d: d[1], reverse=True)[
:50]):
print('{index}:{data}'.format(index=index + 1, data=data))
# 获取番剧合集弹幕排行榜
def count(self, url, desc=None):
bytes = requests.get(url).content
comment_selector = etree.HTML(bytes)
if not desc is None:
print(desc)
print("url=%s" % url)
for comment in comment_selector.xpath('//i//d/text()'):
if comment in self.obj["data"]:
with self.lock:
self.obj["data"][comment] = self.obj["data"][comment] + 1
else:
with self.lock:
self.obj["data"][comment] = 1
if not self.obj["flag"]:
for keyword in self.keywords:
if keyword in comment:
self.obj["flag"] = True
# 根据cid获取历史弹幕地址
def getCidUrls(cid):
urls = []
url = "https://comment.bilibili.com/rolldate,%d" % cid
req = requests.get(url)
if len(req.text) > 0:
for i in json.loads(req.text):
urls.append("https://comment.bilibili.com/dmroll,%s,%d" % (i['timestamp'], cid))
else:
urls.append("https://comment.bilibili.com/%d.xml" % cid)
return urls
# 下载弹幕
def downloadXml(path, cid, size=None, histroy=True):
dlist = set()
flag = None
if histroy:
flag = parseXml(getCidUrls(cid), dlist, size)
else:
parseXml("https://comment.bilibili.com/%d.xml" % cid, dlist, size)
if size is None or (histroy and not size is None and flag):
if os.path.exists(path):
shutil.rmtree(path)
os.makedirs(path)
f = open('{path}/{cid}.xml'.format(path=path, cid=cid), 'wb')
f.write(b'<?xml version="1.0" encoding="UTF-8"?><i>')
for i in dlist:
f.write(('\r\n' + i).encode())
f.write(b'\r\n</i>')
f.close()
def xml(url):
bytes = requests.get(url).content
return etree.HTML(bytes)
def parseXml(urls, dlist, size=None):
if isinstance(urls, str):
urls = [urls]
if not size is None:
size = float(size.strip('%')) / 100.0
for url in urls:
comment_selector = xml(url)
list = comment_selector.xpath('//i//d/text()')
maxlimit = int(comment_selector.xpath('//i//maxlimit/text()')[0])
if len(list) > 0:
print('弹幕数:{list},最大弹幕数:{maxlimit},弹幕池填充:{p}'.format(list=len(list), maxlimit=maxlimit,
p='%.2f%%' % (len(list) / maxlimit * 100)))
for element in comment_selector.xpath('//i//d'):
if len(element.xpath("text()")) > 0:
fstr = '<d p="{p}">{content}</d>'.format(p=str(element.xpath("@p")[0]),
content=str(element.xpath("text()")[0]))
dlist.add(fstr)
currentSize = len(dlist) / maxlimit
print('填充率:{l}'.format(l='%.2f%%' % (currentSize * 100)))
if not size is None and currentSize >= size:
return True
return False
if __name__ == '__main__':
cids = [7636499, 7636501, 7636500, 7636503, 7636504, 7636502, 7636509, 7636508, 7636506, 7636507, 7636505]
downloadXml('F:/ABC',12026697,histroy=False)