From fce829a4e4415b46556708e32fd3507b7fd67a16 Mon Sep 17 00:00:00 2001
From: p6fxi93qh <1240380517@qq.com>
Date: Tue, 17 Sep 2024 23:23:31 +0800
Subject: [PATCH] Update getdm.py
---
getdm.py | 120 +++++++++++++++++++++++++++----------------------------
1 file changed, 58 insertions(+), 62 deletions(-)
diff --git a/getdm.py b/getdm.py
index 92ab467..c7749bc 100644
--- a/getdm.py
+++ b/getdm.py
@@ -1,62 +1,58 @@
-
-import re
-import requests
-from lxml import etree
-import time
-from urllib.parse import quote
-
-headers = {"User-Agent":"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/128.0.0.0 Safari/537.36 Edg/128.0.0.0"}
-
-cookies = " buvid3=8231EFA3-C943-B46F-C28D-10CA6EB40DD180949infoc; CURRENT_FNVAL=4048; rpdid=|(JYYJ|u)luu0J'uYmkl~mY)k; buvid4=CA2A9538-3AB7-5FF0-1AD2-830935987DFB07953-024012212-75GkjfNGeaGAWMbBOvqWtQ%3D%3D; buvid_fp=f29e4e3ef42f7eb5eeb379115e870ad9; LIVE_BUVID=AUTO7217086875647310; PVID=2; b_nut=100; _uuid=7FDC1091C-5D99-24A8-4133-F89DFB510EA6531786infoc; header_theme_version=CLOSE; enable_web_push=DISABLE; bili_ticket=eyJhbGciOiJIUzI1NiIsImtpZCI6InMwMyIsInR5cCI6IkpXVCJ9.eyJleHAiOjE3MjY2NDk2NTcsImlhdCI6MTcyNjM5MDM5NywicGx0IjotMX0.3Z1vsF3Xzu2nPlPrZBCesXytvgQb_37-XNcBtrRlMzE; bili_ticket_expires=1726649597; bsource=search_bing; bmg_af_switch=1; bmg_src_def_domain=i1.hdslb.com; home_feed_column=4; SESSDATA=188e2907%2C1742055275%2C569b4%2A92CjDCOHUaxUkq_L5c6PfUab0-sYNi5nOwFET1z4H9q14EuO2HVsnxLmx1932gpvUQBgESVnFIY3lzUFJhZWxtY1N1d21POEc5bDRKRldVTlJSTDNWdG8ybkJSYmdGN3g3S2hjRk1uYzhSVEJUUkVOMThzb2Zfb0txZlUwbDNVdnZEM1h6UjJ6amhBIIEC; bili_jct=b1a79048046b927643ac32fe2ce260a0; DedeUserID=1986103749; DedeUserID__ckMd5=be6d80b5f24d01e0; browser_resolution=1220-150; sid=6jgymz8o; bp_t_offset_1986103749=977826403892330496; b_lsid=EF11D7E6_191FBE2F27F"
-
-dic_cookies = {}
-for i in cookies.split("; "):
- dic_cookies[i.split("=")[0]] = i.split("=")[1]
-
-#获取整页视频的url
-def get_videos_url(keyword,pages=1):
- url = f"https://search.bilibili.com/video?keyword={keyword}&from_source=webtop_search&spm_id_from=333.1007&search_source=5"
- result = []
- for i in range(0, pages):
- if i > 0 :
- url1 = f"{url}&page={i + 1}&o={i * 30}"
- else:
- url1 = url
- response = requests.get(url1, headers=headers,cookies=dic_cookies)
- response.encoding = 'utf-8'
- e = etree.HTML(response.text)
- # xpath定位
- a = e.xpath('//div[@class="video-list row"]/div[not(contains(@class, "to_hide_md"))]/div[1]/div[2]/a[1]')
- for element in a:
- href = element.get('href') # 获取href属性
- href1 = f"https:{href}"
- result.append(href1)
- print(href1)
- time.sleep(5)
- return result
-
-
-def get_danmu(video_id):
-
- #打开视频页面，获取视频cid
- response = requests.get(video_id, headers=headers)
- video_cid = re.search(r'"cid":(\d*),', response.text).group(1)
-
- #用现有链接获取弹幕
- url_dm = f"https://comment.bilibili.com/{video_cid}.xml"
- response_dm = requests.get(url_dm,headers=headers,cookies=dic_cookies)
- if response.status_code == 200:
- response_dm.encoding = 'utf-8'
- content = re.findall('(.*?)', response_dm.text)
- print("请求弹幕成功")
- return content
- else:
- print("请求弹幕失败")
- return
-
-
-def writeintxt (list,file_path):
- with open(file_path, 'w', encoding='utf-8') as file:
- for item in list:
- file.write("%s\n" % item)
-
+import re
+import requests
+from lxml import etree
+import time
+from urllib.parse import quote
+
+headers = {"User-Agent":"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/128.0.0.0 Safari/537.36 Edg/128.0.0.0"}
+
+cookies = " buvid3=8231EFA3-C943-B46F-C28D-10CA6EB40DD180949infoc; CURRENT_FNVAL=4048; rpdid=|(JYYJ|u)luu0J'uYmkl~mY)k; buvid4=CA2A9538-3AB7-5FF0-1AD2-830935987DFB07953-024012212-75GkjfNGeaGAWMbBOvqWtQ%3D%3D; buvid_fp=f29e4e3ef42f7eb5eeb379115e870ad9; LIVE_BUVID=AUTO7217086875647310; PVID=2; b_nut=100; _uuid=7FDC1091C-5D99-24A8-4133-F89DFB510EA6531786infoc; header_theme_version=CLOSE; enable_web_push=DISABLE; bili_ticket=eyJhbGciOiJIUzI1NiIsImtpZCI6InMwMyIsInR5cCI6IkpXVCJ9.eyJleHAiOjE3MjY2NDk2NTcsImlhdCI6MTcyNjM5MDM5NywicGx0IjotMX0.3Z1vsF3Xzu2nPlPrZBCesXytvgQb_37-XNcBtrRlMzE; bili_ticket_expires=1726649597; bsource=search_bing; bmg_af_switch=1; bmg_src_def_domain=i1.hdslb.com; home_feed_column=4; SESSDATA=188e2907%2C1742055275%2C569b4%2A92CjDCOHUaxUkq_L5c6PfUab0-sYNi5nOwFET1z4H9q14EuO2HVsnxLmx1932gpvUQBgESVnFIY3lzUFJhZWxtY1N1d21POEc5bDRKRldVTlJSTDNWdG8ybkJSYmdGN3g3S2hjRk1uYzhSVEJUUkVOMThzb2Zfb0txZlUwbDNVdnZEM1h6UjJ6amhBIIEC; bili_jct=b1a79048046b927643ac32fe2ce260a0; DedeUserID=1986103749; DedeUserID__ckMd5=be6d80b5f24d01e0; browser_resolution=1220-150; sid=6jgymz8o; bp_t_offset_1986103749=977826403892330496; b_lsid=EF11D7E6_191FBE2F27F"
+
+dic_cookies = {}
+for i in cookies.split("; "):
+ dic_cookies[i.split("=")[0]] = i.split("=")[1]
+
+#获取整页视频的url
+def get_videos_url(keyword,pages=1):
+ url = f"https://search.bilibili.com/video?keyword={keyword}&from_source=webtop_search&spm_id_from=333.1007&search_source=5"
+ result = []
+ for i in range(0, pages):
+ if i > 0 :
+ url1 = f"{url}&page={i + 1}&o={i * 30}"
+ else:
+ url1 = url
+ response = requests.get(url1, headers=headers,cookies=dic_cookies)
+ response.encoding = 'utf-8'
+ e = etree.HTML(response.text)
+ # xpath定位
+ a = e.xpath('//div[@class="video-list row"]/div[not(contains(@class, "to_hide_md"))]/div[1]/div[2]/a[1]')
+ for element in a:
+ href = element.get('href') # 获取href属性
+ href1 = f"https:{href}"
+ result.append(href1)
+ print(href1)
+ time.sleep(5)
+ return result
+
+
+def get_danmu(video_id):
+ #打开视频页面，获取视频cid
+ response = requests.get(video_id, headers=headers)
+ video_cid = re.search(r'"cid":(\d*),', response.text).group(1)
+
+ #用现有链接获取弹幕
+ url_dm = f"https://comment.bilibili.com/{video_cid}.xml"
+ response_dm = requests.get(url_dm,headers=headers,cookies=dic_cookies)
+ if response.status_code == 200:
+ response_dm.encoding = 'utf-8'
+ content = re.findall('(.*?)', response_dm.text)
+ print("请求弹幕成功")
+ return content
+ else:
+ print("请求弹幕失败")
+ return
+
+def writeintxt (list,file_path):
+ with open(file_path, 'w', encoding='utf-8') as file:
+ for item in list:
+ file.write("%s\n" % item)
\ No newline at end of file