增加优酷弹幕下载以及其他改进

.gitignore

@@ -1,6 +1,7 @@
 # 额外
 .vscode/
 releases/
+test/
 *.ass
 methods/calc_danmu_pos.py
 

GetDanMu.py

@@ -3,7 +3,7 @@
 '''
 # 作者: weimo
 # 创建日期: 2020-01-04 19:14:39
-# 上次编辑时间: 2020-01-05 14:47:53
+# 上次编辑时间       : 2020-01-11 17:49:40
 # 一个人的命运啊,当然要靠自我奋斗,但是...
 '''
 
@@ -13,6 +13,7 @@ from argparse import ArgumentParser
 
 from sites.qq import main as qq
 from sites.iqiyi import main as iqiyi
+from sites.youku import main as youku
 from pfunc.cfunc import check_url_site
 
 # -------------------------------------------
@@ -34,8 +35,9 @@ def main():
     parser.add_argument("-vid", "--vid", default="", help="下载vid对应视频的弹幕，支持同时多个vid，需要用逗号隔开")
     parser.add_argument("-aid", "--aid", default="", help="下载aid对应视频的弹幕（爱奇艺合集）")
     parser.add_argument("-tvid", "--tvid", default="", help="下载tvid对应视频的弹幕，支持同时多个tvid，需要用逗号隔开")
+    parser.add_argument("-series", "--series", action="store_true", help="尝试通过单集得到合集的全部弹幕")
     parser.add_argument("-u", "--url", default="", help="下载视频链接所指向视频的弹幕")
-    parser.add_argument("-y", "--y", action="store_true", help="覆盖原有弹幕而不提示")
+    parser.add_argument("-y", "--y", action="store_false", help="默认覆盖原有弹幕而不提示")
     args = parser.parse_args()
     # print(args.__dict__)
     if args.url != "":
@@ -46,6 +48,8 @@ def main():
         subtitles = qq(args)
     if args.site == "iqiyi":
         subtitles = iqiyi(args)
+    if args.site == "youku":
+        subtitles = youku(args)
 
 if __name__ == "__main__":
     # 打包 --> pyinstaller GetDanMu.spec

README.md

@@ -1,7 +1,7 @@
 <!--
  * @作者: weimo
  * @创建日期: 2020-01-04 18:45:58
- * @上次编辑时间       : 2020-01-05 14:42:42
+ * @上次编辑时间       : 2020-01-11 17:48:19
  * @一个人的命运啊,当然要靠自我奋斗,但是...
  -->
 # GetDanMu
@@ -13,6 +13,7 @@
 | :--: | :-- | :-----: | :-----: | :-----: |
 | **腾讯视频** | <https://v.qq.com/>    |✓|✓| |
 | **爱奇艺** | <https://www.iqiyi.com/>    |✓|✓|✓|
+| **优酷** | <https://v.youku.com/>    |✓|✓|✓|
 
 ## 可能存在的问题
 - 下载进度接近100%时暂时没有反应
@@ -24,6 +25,12 @@
 
 # 更新日志
 
+## 2020/1/11
+- 增加优酷弹幕下载，支持合集，支持通过单集直接下载合集弹幕（暂时仅限优酷）
+- 改进去重方式
+- 优酷的视频id用vid指代，若下载合集请使用连接或通过`-series`选项下载合集弹幕
+- 加入下载进度显示，后续可能改进为单行刷新
+
 ## 2020/1/5
 
 - 增加了通过链接下载爱奇艺视频弹幕的方法，支持综艺合集。

basic/ass.py

@@ -3,7 +3,7 @@
 '''
 # 作者: weimo
 # 创建日期: 2020-01-04 19:14:46
-# 上次编辑时间: 2020-01-05 14:45:55
+# 上次编辑时间       : 2020-01-11 17:20:21
 # 一个人的命运啊,当然要靠自我奋斗,但是...
 '''
 
@@ -62,10 +62,4 @@ def check_font(font):
         font_style_name = font
     else:
         pass
-    return font_path, font_style_name
+    return font_path, font_style_name
-
-def check_content(content: str, comments: list):
-    content = content.replace(" ", "")
-    if content in comments:
-        return
-    return content

pfunc/cfunc.py

@@ -3,11 +3,24 @@
 '''
 # 作者: weimo
 # 创建日期: 2020-01-05 12:45:18
-# 上次编辑时间: 2020-01-05 14:44:42
+# 上次编辑时间       : 2020-01-11 17:37:22
 # 一个人的命运啊,当然要靠自我奋斗,但是...
 '''
+
+import hashlib
 from urllib.parse import urlparse
 
+def remove_same_danmu(comments: list):
+    # 在原有基础上pop会引起索引变化 所以还是采用下面这个方式
+    contents = []
+    for comment in comments:
+        content, color, timepoint = comment
+        content = content.replace(" ", "")
+        if content in contents:
+            continue
+        else:
+            contents.append([content, color, timepoint])
+    return contents
 
 def check_url_site(url):
     return urlparse(url).netloc.split(".")[-2]
@@ -21,4 +34,11 @@ def check_url_locale(url):
     if urlparse(url).netloc.split(".")[0] == "tw":
         return flag["tw"]
     else:
-        return flag["cn"] 
+        return flag["cn"] 
+
+def yk_msg_sign(msg: str):
+    return hashlib.new("md5", bytes(msg + "MkmC9SoIw6xCkSKHhJ7b5D2r51kBiREr", "utf-8")).hexdigest()
+
+def yk_t_sign(token, t, appkey, data):
+    text = "&".join([token, t, appkey, data])
+    return hashlib.new('md5', bytes(text, 'utf-8')).hexdigest()

pfunc/dump_to_ass.py

@@ -3,7 +3,7 @@
 '''
 # 作者: weimo
 # 创建日期: 2020-01-04 19:17:44
-# 上次编辑时间: 2020-01-05 14:45:03
+# 上次编辑时间       : 2020-01-11 17:25:09
 # 一个人的命运啊,当然要靠自我奋斗,但是...
 '''
 import os
@@ -11,16 +11,19 @@ import os
 from basic.ass import get_ass_head, check_font
 from methods.assbase import ASS
 from methods.sameheight import SameHeight
+from pfunc.cfunc import remove_same_danmu
 
 def write_one_video_subtitles(file_path, comments, args):
     # 对于合集则每次都都得检查一次 也可以放在上一级 放在这里 考虑后面可能特殊指定字体的情况
     font_path, font_style_name = check_font(args.font)
     ass_head = get_ass_head(font_style_name, args.font_size)
-    get_xy_obj = SameHeight("那就写这一句作为初始化测试吧！", font_path=font_path, font_size=args.font_size)
+    get_xy_obj = SameHeight("那就写这一句作为初始化测试吧！", font_path=font_path, font_size=int(args.font_size))
     subtitle = ASS(file_path, get_xy_obj, font=font_style_name)
+    comments = remove_same_danmu(comments)
     for comment in comments:
         subtitle.create_new_line(comment)
     write_lines_to_file(ass_head, subtitle.lines, file_path)
+    return comments
 
 def write_lines_to_file(ass_head, lines, file_path):
     with open(file_path, "a+", encoding="utf-8") as f:

pfunc/request_info.py

@@ -3,7 +3,7 @@
 '''
 # 作者: weimo
 # 创建日期: 2020-01-04 19:14:43
-# 上次编辑时间: 2020-01-05 14:47:16
+# 上次编辑时间       : 2020-01-11 17:42:30
 # 一个人的命运啊,当然要靠自我奋斗,但是...
 '''
 import re
@@ -192,4 +192,64 @@ def get_vinfos_by_url(url):
         else:
             return get_vinfos(aid, locale=locale)
 
-#-------------------------------------------iqiyi--------------------------------------------
+#-------------------------------------------iqiyi--------------------------------------------
+
+#-------------------------------------------youku--------------------------------------------
+
+def get_vinfos_by_url_youku(url, isall=False):
+    vid_patterns = ["[\s\S]+?youku.com/video/id_(/+?)\.html", "[\s\S]+?youku.com/v_show/id_(.+?)\.html"]
+    video_id = matchit(vid_patterns, url)
+    show_id_patterns = ["[\s\S]+?youku.com/v_nextstage/id_(/+?)\.html", "[\s\S]+?youku.com/show/id_z(.+?)\.html", "[\s\S]+?youku.com/show_page/id_z(.+?)\.html", "[\s\S]+?youku.com/alipay_video/id_(.+?)\.html"]
+    show_id = matchit(show_id_patterns, url)
+    if video_id is None and show_id is None:
+        return None
+    if video_id:
+        return get_vinfos_by_video_id(video_id, isall=isall)
+    if show_id.__len__() == 20 and show_id == show_id.lower():
+        return get_vinfos_by_show_id(show_id)
+    else:
+        return get_vinfos_by_video_id(show_id, isall=isall)
+
+def get_vinfos_by_video_id(video_id, isall=False):
+    api_url = "https://openapi.youku.com/v2/videos/show.json?client_id=53e6cc67237fc59a&package=com.huawei.hwvplayer.youku&ext=show&video_id={}".format(video_id)
+    try:
+        r = requests.get(api_url, headers=chrome, timeout=5).content.decode("utf-8")
+    except Exception as e:
+        print("get_vinfos_by_video_id error info -->", e)
+        return None
+    data = json.loads(r)
+    if isall:
+        show_id = data["show"]["id"]
+        return get_vinfos_by_show_id(show_id)
+    duration = 0
+    if data.get("duration"):
+        duration = int(float(data["duration"]))
+    if data.get("title"):
+        name = data["title"] + "_" + str(duration)
+    else:
+        name = "优酷未知" + "_" + str(duration)
+    vinfo = [name, duration, video_id]
+    return [vinfo]
+
+def get_vinfos_by_show_id(show_id):
+    api_url = "https://openapi.youku.com/v2/shows/videos.json?show_videotype=正片&count=100&client_id=53e6cc67237fc59a&page=1&show_id={}&package=com.huawei.hwvplayer.youku".format(show_id)
+    try:
+        r = requests.get(api_url, headers=chrome, timeout=5).content.decode("utf-8")
+    except Exception as e:
+        print("get_vinfos_by_show_id error info -->", e)
+        return None
+    data = json.loads(r)["videos"]
+    if data.__len__() == 0:
+        return None
+    vinfos = []
+    for video in data:
+        duration = 0
+        if video.get("duration"):
+            duration = int(float(video["duration"]))
+        if video.get("title"):
+            name = video["title"] + "_" + str(duration)
+        else:
+            name = "优酷未知_{}".format(video["id"]) + "_" + str(duration)
+        vinfos.append([name, duration, video["id"]])
+    return vinfos
+#-------------------------------------------youku--------------------------------------------

sites/iqiyi.py

@@ -3,7 +3,7 @@
 '''
 # 作者: weimo
 # 创建日期: 2020-01-04 19:14:41
-# 上次编辑时间: 2020-01-05 14:45:17
+# 上次编辑时间       : 2020-01-11 17:23:32
 # 一个人的命运啊,当然要靠自我奋斗,但是...
 '''
 
@@ -14,7 +14,6 @@ from zlib import decompress
 from xmltodict import parse
 
 from basic.vars import iqiyiplayer
-from basic.ass import check_content
 from pfunc.dump_to_ass import check_file, write_one_video_subtitles
 from pfunc.request_info import get_vinfos, get_vinfos_by_url, get_vinfo_by_tvid
 
@@ -48,7 +47,6 @@ def get_danmu_by_tvid(name, duration, tvid):
             continue
         # with open("raw_xml.json", "w", encoding="utf-8") as f:
         #     f.write(json.dumps(parse(raw_xml), ensure_ascii=False, indent=4))
-        contents = []
         if entry.__class__ != list:
             entry = [entry]
         for comment in entry:
@@ -58,13 +56,8 @@ def get_danmu_by_tvid(name, duration, tvid):
             if bulletInfo.__class__ != list:
                 bulletInfo = [bulletInfo]
             for info in bulletInfo:
-                content = check_content(info["content"], contents)
-                if content is None:
-                    continue
-                else:
-                    contents.append(content)
                 color = [info["color"]]
-                comments.append([content, color, int(comment["int"])])
+                comments.append([info["content"], color, int(comment["int"])])
         print("已下载{:.2f}%".format(index * timestamp * 100 / duration))
         index += 1
     comments = sorted(comments, key=lambda _: _[-1])
@@ -93,8 +86,8 @@ def main(args):
         flag, file_path = check_file(name, skip=args.y)
         if flag is False:
             print("跳过{}".format(name))
-            return
+            continue
         comments = get_danmu_by_tvid(name, duration, tvid)
-        write_one_video_subtitles(file_path, comments, args)
+        comments = write_one_video_subtitles(file_path, comments, args)
         subtitles.update({file_path:comments})
     return subtitles

sites/qq.py

@@ -3,7 +3,7 @@
 '''
 # 作者: weimo
 # 创建日期: 2020-01-04 19:14:37
-# 上次编辑时间: 2020-01-05 14:47:36
+# 上次编辑时间       : 2020-01-11 17:25:34
 # 一个人的命运啊,当然要靠自我奋斗,但是...
 '''
 
@@ -13,7 +13,6 @@ import json
 import requests
 
 from basic.vars import qqlive
-from basic.ass import check_content
 from pfunc.dump_to_ass import check_file, write_one_video_subtitles
 from pfunc.request_info import get_all_vids_by_cid as get_vids
 from pfunc.request_info import get_danmu_target_id_by_vid as get_target_id
@@ -80,13 +79,7 @@ def get_danmu_by_target_id(vid: str, duration: int, target_id, font="微软雅
             # timestamp不变 再试一次
             continue
         danmu_count = danmu["count"]
-        contents = []
         for comment in danmu["comments"]:
-            content = check_content(comment["content"], contents)
-            if content is None:
-                continue
-            else:
-                contents.append(content)
             if comment["content_style"]:
                 style = json.loads(comment["content_style"])
                 if style.get("gradient_colors"):
@@ -97,7 +90,7 @@ def get_danmu_by_target_id(vid: str, duration: int, target_id, font="微软雅
                     color = ["ffffff"]
             else:
                 color = ["ffffff"]
-            comments.append([content, color, comment["timepoint"]])
+            comments.append([comment["content"], color, comment["timepoint"]])
         print("已下载{:.2f}%".format(params["timestamp"]*100/duration))
         params["timestamp"] += 30
     comments = sorted(comments, key=lambda _: _[-1])
@@ -157,8 +150,11 @@ def main(args):
     vinfos = get_video_info_by_vid(vids)
     subtitles = {}
     for vinfo in vinfos:
-        comments, file_path = get_one_subtitle_by_vinfo(vinfo, args.font, args.font_size, args.y)
+        infos = get_one_subtitle_by_vinfo(vinfo, args.font, args.font_size, args.y)
-        write_one_video_subtitles(file_path, comments, args)
+        if infos is None:
+            continue
+        comments, file_path = infos
+        comments = write_one_video_subtitles(file_path, comments, args)
         subtitles.update({file_path:comments})
     return subtitles
 

sites/youku.py

@@ -0,0 +1,141 @@
+#!/usr/bin/env python3.7
+# coding=utf-8
+'''
+# 作者: weimo
+# 创建日期: 2020-01-05 14:52:21
+# 上次编辑时间       : 2020-01-11 17:53:14
+# 一个人的命运啊,当然要靠自我奋斗,但是...
+'''
+import re
+import time
+import json
+import base64
+import requests
+
+from basic.vars import chrome
+from pfunc.dump_to_ass import check_file, write_one_video_subtitles
+from pfunc.cfunc import yk_msg_sign, yk_t_sign
+from pfunc.request_info import get_vinfos_by_show_id, get_vinfos_by_video_id, get_vinfos_by_url_youku
+
+def get_tk_enc():
+    """
+    获取优酷的_m_h5_tk和_m_h5_tk_enc
+    """
+    api_url = "https://acs.youku.com/h5/mtop.com.youku.aplatform.weakget/1.0/?jsv=2.5.1&appKey=24679788"
+    try:
+        r = requests.get(api_url, headers=chrome, timeout=5)
+    except Exception as e:
+        return
+    tk_enc = dict(r.cookies)
+    if tk_enc.get("_m_h5_tk_enc") and tk_enc.get("_m_h5_tk"):
+        return tk_enc
+    return
+
+def get_cna():
+    api_url = "https://log.mmstat.com/eg.js"
+    try:
+        r = requests.get(api_url, headers=chrome, timeout=5)
+    except Exception as e:
+        return
+    cookies = dict(r.cookies)
+    if cookies.get("cna"):
+        return cookies["cna"]
+    return
+
+def get_danmu_by_mat(vid, cna, mat: int, comments: list):
+    api_url = "https://acs.youku.com/h5/mopen.youku.danmu.list/1.0/"
+    tm = str(int(time.time() * 1000))
+    msg = {
+        "ctime": tm,
+        "ctype": 10004,
+        "cver": "v1.0",
+        "guid": cna,
+        "mat": mat,
+        "mcount": 1,
+        "pid": 0,
+        "sver": "3.1.0",
+        "type": 1,
+        "vid": vid}
+    msg_b64encode = base64.b64encode(json.dumps(msg, separators=(',', ':')).encode("utf-8")).decode("utf-8")
+    msg.update({"msg":msg_b64encode})
+    msg.update({"sign":yk_msg_sign(msg_b64encode)})
+    # 测试发现只要有Cookie的_m_h5_tk和_m_h5_tk_enc就行
+    tk_enc = get_tk_enc()
+    if tk_enc is None:
+        return
+    headers = {
+        "Content-Type":"application/x-www-form-urlencoded",
+        "Cookie":";".join([k + "=" + v for k, v in tk_enc.items()]),
+        "Referer": "https://v.youku.com"
+    }
+    headers.update(chrome)
+    t = str(int(time.time() * 1000))
+    data = json.dumps(msg, separators=(',', ':'))
+    params = {
+        "jsv":"2.5.6",
+        "appKey":"24679788",
+        "t":t,
+        "sign":yk_t_sign(tk_enc["_m_h5_tk"][:32], t, "24679788", data),
+        "api":"mopen.youku.danmu.list",
+        "v":"1.0",
+        "type":"originaljson",
+        "dataType":"jsonp",
+        "timeout":"20000",
+        "jsonpIncPrefix":"utility"
+    }
+    try:
+        r = requests.post(api_url, params=params, data={"data":data}, headers=headers, timeout=5).content.decode("utf-8")
+    except Exception as e:
+        print("youku danmu request failed.", e)
+        return "once again"
+    result = json.loads(json.loads(r)["data"]["result"])["data"]["result"]
+    for item in result:
+        comment = item["content"]
+        c_int = json.loads(item["propertis"])["color"]
+        if c_int.__class__ == str:
+            c_int = int(c_int)
+        color = hex(c_int)[2:].zfill(6)
+        timepoint = item["playat"] / 1000
+        comments.append([comment, [color], timepoint])
+    return comments
+
+def main(args):
+    cna = get_cna()
+    if cna is None:
+        # 放前面 免得做无用功
+        return
+    isall = False
+    if args.series:
+        isall = True
+    vinfos = []
+    if args.url:
+        vi = get_vinfos_by_url_youku(args.url, isall=isall)
+        if vi:
+            vinfos += vi
+    if args.vid:
+        vi = get_vinfos_by_video_id(args.vid, isall=isall)
+        if vi:
+            vinfos += vi
+    subtitles = {}
+    for name, duration, video_id in vinfos:
+        print(name, "开始下载...")
+        flag, file_path = check_file(name, skip=args.y)
+        if flag is False:
+            print("跳过{}".format(name))
+            continue
+        max_mat = duration // 60 + 1
+        comments = []
+        for mat in range(max_mat):
+            result = get_danmu_by_mat(video_id, cna, mat + 1, comments)
+            if result is None:
+                continue
+            elif result == "once again":
+                # 可能改成while好点
+                result = get_danmu_by_mat(video_id, cna, mat + 1, comments)
+                if result is None:
+                    continue
+            comments = result
+            print("已下载{}/{}".format(mat + 1, max_mat))
+        comments = write_one_video_subtitles(file_path, comments, args)
+        subtitles.update({file_path:comments})
+    return subtitles