Compare commits

...

36 Commits

Author SHA1 Message Date
taizan-hokuto
ac0f052aa0 Merge branch 'hotfix' 2019-12-30 19:11:39 +09:00
taizan-hokuto
1cc0338a8e Export DefaultProcessor 2019-12-30 19:09:12 +09:00
taizan-hokuto
f6b8229998 Merge branch 'release' 2019-12-30 18:33:41 +09:00
taizan-hokuto
f24c5f9e30 Increment version 2019-12-30 18:32:17 +09:00
taizan-hokuto
5268961854 Delete unnecessary lines of old logger 2019-12-30 17:54:53 +09:00
taizan-hokuto
733f754e11 Merge branch 'feature/fix_replaychat_is_alive' into develop 2019-12-30 17:32:50 +09:00
taizan-hokuto
582d0b749d Add comment 2019-12-30 17:32:23 +09:00
taizan-hokuto
b8bc00d880 Fix README 2019-12-30 17:27:25 +09:00
taizan-hokuto
ce96d94e23 Fix termination of fetching chat data 2019-12-30 17:10:34 +09:00
taizan-hokuto
7af92f14c0 Merge branch 'feature/config_logger' into develop 2019-12-30 15:01:32 +09:00
taizan-hokuto
7305e4178b Change description of getting logger 2019-12-30 14:38:02 +09:00
taizan-hokuto
a835d58e10 Merge branch 'feature/combinator' into develop 2019-12-30 11:04:25 +09:00
taizan-hokuto
4e956b8d84 Implement Combinator, DummyProcessor 2019-12-30 11:02:29 +09:00
taizan-hokuto
c4f1194a53 Merge branch 'feature/fix_massage_ex' into develop 2019-12-30 10:36:22 +09:00
taizan-hokuto
90b10a9f8f Integrate rendering message and message_ex 2019-12-27 02:19:08 +09:00
taizan-hokuto
b576c3f928 Merge branch 'develop' 2019-12-24 00:57:06 +09:00
taizan-hokuto
c0728e1366 Increment version 2019-12-24 00:55:00 +09:00
taizan-hokuto
fff09d4c27 Fix README 2019-12-24 00:52:53 +09:00
taizan-hokuto
810b6c8c6b Fix README 2019-12-24 00:29:10 +09:00
taizan-hokuto
dfada86caf Merge branch 'develop' 2019-12-22 02:56:54 +09:00
taizan-hokuto
91aa944df5 Increment version 2019-12-22 02:54:20 +09:00
taizan-hokuto
6ac5191e85 Change debug mode 2019-12-22 01:58:42 +09:00
taizan-hokuto
fff3e0371f Export SpeedCalculator 2019-12-22 01:39:35 +09:00
taizan-hokuto
a70efe8a67 Fix comments 2019-12-21 23:42:51 +09:00
taizan-hokuto
dc47f4debe Fix README 2019-12-21 22:40:08 +09:00
taizan-hokuto
ab5a2a8df2 Fix syntax error 2019-12-21 20:12:30 +09:00
taizan-hokuto
5a79f26fa7 Fix calculation algorithm 2019-12-21 20:06:55 +09:00
taizan-hokuto
18666199b7 Add test SpeedCalculator 2019-12-21 02:13:15 +09:00
taizan-hokuto
b357bccb98 Add test json 2019-12-20 23:57:10 +09:00
taizan-hokuto
3c1f079d5f Extends ChatProcessor explicitly 2019-12-20 21:33:32 +09:00
taizan-hokuto
289841a000 Export api speed_calculator 2019-12-20 01:13:50 +09:00
taizan-hokuto
ee0ff7fe74 Merge branch 'develop' 2019-12-19 23:14:37 +09:00
taizan-hokuto
c0870ce537 Fix README 2019-12-19 23:14:05 +09:00
taizan-hokuto
de6ef2490e Merge branch 'develop' 2019-12-19 02:03:09 +09:00
taizan-hokuto
b8bdbdc36f Increment version 2019-12-19 02:02:32 +09:00
taizan-hokuto
9f5d3f323e Fix README 2019-12-19 02:00:41 +09:00
25 changed files with 788 additions and 188 deletions

View File

@@ -8,7 +8,7 @@ pytchat is a python library for fetching youtube live chat
without using youtube api, Selenium or BeautifulSoup.
Other features:
+ Customizable chat data processors including yt api compatible one.
+ Customizable chat data processors including youtube api compatible one.
+ Available on asyncio context.
+ Quick fetching of initial chat data by generating continuation params
instead of web scraping.
@@ -27,7 +27,7 @@ pip install pytchat
```python
from pytchat import LiveChat
chat = LiveChat("G1w62uEMZ74")
chat = LiveChat("rsHWP7IjMiw")
while chat.is_alive():
data = chat.get()
for c in data.items:
@@ -40,17 +40,17 @@ while chat.is_alive():
from pytchat import LiveChat
import time
def main()
chat = LiveChat("G1w62uEMZ74", callback = func)
while chat.is_alive():
time.sleep(3)
#other background operation.
#callback function is automatically called periodically.
def func(data):
#callback function is automatically called.
def display(data):
for c in data.items:
print(f"{c.datetime} [{c.author.name}]-{c.message} {c.amountString}")
data.tick()
#entry point
chat = LiveChat("rsHWP7IjMiw", callback = display)
while chat.is_alive():
time.sleep(3)
#other background operation.
```
### asyncio context:
@@ -59,61 +59,57 @@ from pytchat import LiveChatAsync
import asyncio
async def main():
chat = LiveChatAsync("G1w62uEMZ74", callback = func)
chat = LiveChatAsync("rsHWP7IjMiw", callback = func)
while chat.is_alive():
await asyncio.sleep(3)
#other background operation.
#callback function is automatically called periodically.
#callback function is automatically called.
async def func(data):
for c in data.items:
print(f"{c.datetime} [{c.author.name}]-{c.message} {c.amountString}")
await data.tick_async()
try:
loop = asyncio.get_event_loop()
loop.run_until_complete(main())
except CancelledError:
pass
```
### youtube api compatible processor:
```python
from pytchat import LiveChat, CompatibleProcessor
import time
chat = LiveChat("G1w62uEMZ74",
chat = LiveChat("rsHWP7IjMiw",
processor = CompatibleProcessor() )
while chat.is_alive():
data = chat.get()
polling = data["pollingIntervalMillis"]/1000
for c in data["items"]:
if c.get("snippet"):
polling = data['pollingIntervalMillis']/1000
for c in data['items']:
if c.get('snippet'):
print(f"[{c['authorDetails']['displayName']}]"
f"-{c['snippet']['displayMessage']}")
time.sleep(polling/len(data["items"]))
time.sleep(polling/len(data['items']))
```
### replay:
```python
from pytchat import ReplayChatAsync
import asyncio
from pytchat import ReplayChat
async def main():
chat = ReplayChatAsync("G1w62uEMZ74", seektime = 1000, callback = func)
#other background operation here.
def main():
#seektime (seconds): start position of chat.
chat = ReplayChat("ojes5ULOqhc", seektime = 60*30)
while chat.is_alive():
data = chat.get()
for c in data.items:
print(f"{c.elapsedTime} [{c.author.name}]-{c.message} {c.amountString}")
data.tick()
#callback function is automatically called periodically.
async def func(data):
for count in range(0,len(data.items)):
c= data.items[count]
if count!=len(data.items):
tick=data.items[count+1].timestamp -data.items[count].timestamp
else:
tick=0
print(f"<{c.elapsedTime}> [{c.author.name}]-{c.message} {c.amountString}")
await asyncio.sleep(tick/1000)
loop = asyncio.get_event_loop()
loop.run_until_complete(main())
main()
```
## Structure of Default Processor
@@ -152,26 +148,26 @@ Each item can be got with items() function.
<tr>
<td>datetime</td>
<td>str</td>
<td>ex. "2019-10-10 12:34:56"</td>
<td>e.g. "2019-10-10 12:34:56"</td>
</tr>
<td>elapsedTime</td>
<td>str</td>
<td>elapsed time. (ex. "1:02:27") *Replay Only.</td>
<td>elapsed time. (e.g. "1:02:27") *Replay Only.</td>
</tr>
<tr>
<td>amountValue</td>
<td>float</td>
<td>ex. 1,234.0</td>
<td>e.g. 1,234.0</td>
</tr>
<tr>
<td>amountString</td>
<td>str</td>
<td>ex. "$ 1,234"</td>
<td>e.g. "$ 1,234"</td>
</tr>
<tr>
<td>currency</td>
<td>str</td>
<td><a href="https://en.wikipedia.org/wiki/ISO_4217">ISO 4217 currency codes</a> (ex. "USD")</td>
<td><a href="https://en.wikipedia.org/wiki/ISO_4217">ISO 4217 currency codes</a> (e.g. "USD")</td>
</tr>
<tr>
<td>bgColor</td>
@@ -200,7 +196,7 @@ Structure of author object.
<tr>
<td>channelId</td>
<td>str</td>
<td>*chatter's channel ID. NOT broadcasting video's channel ID.</td>
<td>*chatter's channel ID.</td>
</tr>
<tr>
<td>channelUrl</td>

View File

@@ -2,7 +2,7 @@
pytchat is a python library for fetching youtube live chat without using yt api, Selenium, or BeautifulSoup.
"""
__copyright__ = 'Copyright (C) 2019 taizan-hokuto'
__version__ = '0.0.3.5'
__version__ = '0.0.4.1'
__license__ = 'MIT'
__author__ = 'taizan-hokuto'
__author_email__ = '55448286+taizan-hokuto@users.noreply.github.com'
@@ -11,12 +11,16 @@ __url__ = 'https://github.com/taizan-hokuto/pytchat'
__all__ = ["core_async","core_multithread","processors"]
from .api import (
config,
LiveChat,
LiveChatAsync,
ReplayChat,
ReplayChatAsync,
ChatProcessor,
CompatibleProcessor,
DefaultProcessor,
SimpleDisplayProcessor,
JsonfileArchiveProcessor
JsonfileArchiveProcessor,
SpeedCalculator,
DummyProcessor
)

View File

@@ -7,4 +7,6 @@ from .processors.default.processor import DefaultProcessor
from .processors.compatible.processor import CompatibleProcessor
from .processors.simple_display_processor import SimpleDisplayProcessor
from .processors.jsonfile_archive_processor import JsonfileArchiveProcessor
from .processors.speed_calculator import SpeedCalculator
from .processors.dummy_processor import DummyProcessor
from . import config

View File

@@ -1,4 +1,13 @@
import logging
LOGGER_MODE = logging.ERROR
from . import mylogger
LOGGER_MODE = None
headers = {
'user-agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/77.0.3865.120 Safari/537.36'}
def logger(module_name: str):
module_logger = mylogger.get_logger(module_name, mode = LOGGER_MODE)
return module_logger

View File

@@ -0,0 +1,32 @@
from logging import NullHandler, getLogger, StreamHandler, FileHandler, Formatter
import logging
import datetime
def get_logger(modname,mode=logging.DEBUG):
logger = getLogger(modname)
if mode == None:
logger.addHandler(NullHandler())
return logger
logger.setLevel(mode)
#create handler1 for showing info
handler1 = StreamHandler()
my_formatter = MyFormatter()
handler1.setFormatter(my_formatter)
handler1.setLevel(mode)
logger.addHandler(handler1)
#create handler2 for recording log file
if mode <= logging.DEBUG:
handler2 = FileHandler(filename="log.txt", encoding='utf-8')
handler2.setLevel(logging.ERROR)
handler2.setFormatter(my_formatter)
logger.addHandler(handler2)
return logger
class MyFormatter(logging.Formatter):
def format(self, record):
s =(datetime.datetime.fromtimestamp(record.created)).strftime("%m-%d %H:%M:%S")+'| '+ (record.module).ljust(15)+(' { '+record.funcName).ljust(20) +":"+str(record.lineno).rjust(4)+'} - '+record.getMessage()
return s

View File

@@ -11,15 +11,14 @@ from concurrent.futures import CancelledError
from .buffer import Buffer
from ..parser.live import Parser
from .. import config
from .. import mylogger
from ..exceptions import ChatParseException,IllegalFunctionCall
from ..paramgen import liveparam
from ..processors.default.processor import DefaultProcessor
from ..processors.combinator import Combinator
logger = mylogger.get_logger(__name__,mode=config.LOGGER_MODE)
MAX_RETRY = 10
logger = config.logger(__name__)
headers = config.headers
MAX_RETRY = 10
class LiveChatAsync:
@@ -72,6 +71,9 @@ class LiveChatAsync:
exception_handler = None,
direct_mode = False):
self.video_id = video_id
if isinstance(processor, tuple):
self.processor = Combinator(processor)
else:
self.processor = processor
self._buffer = buffer
self._callback = callback
@@ -148,7 +150,7 @@ class LiveChatAsync:
async def _listen(self, continuation):
''' continuationに紐付いたチャットデータを取得し
チャットデータを格納、
Bufferにチャットデータを格納、
次のcontinuaitonを取得してループする。
Parameter
@@ -180,9 +182,11 @@ class LiveChatAsync:
await asyncio.sleep(diff_time)
continuation = metadata.get('continuation')
except ChatParseException as e:
logger.info(f"{str(e)}video_id:\"{self.video_id}\"")
self.terminate()
logger.error(f"{str(e)}video_id:\"{self.video_id}\"")
return
except (TypeError , json.JSONDecodeError) :
self.terminate()
logger.error(f"{traceback.format_exc(limit = -1)}")
return
@@ -211,6 +215,7 @@ class LiveChatAsync:
else:
logger.error(f"[{self.video_id}]"
f"Exceeded retry count. status_code={status_code}")
self.terminate()
return None
return livechat_json

View File

@@ -12,12 +12,12 @@ from queue import Queue
from .buffer import Buffer
from ..parser.replay import Parser
from .. import config
from .. import mylogger
from ..exceptions import ChatParseException,IllegalFunctionCall
from ..paramgen import arcparam
from ..processors.default.processor import DefaultProcessor
from ..processors.combinator import Combinator
logger = mylogger.get_logger(__name__,mode=config.LOGGER_MODE)
logger = config.logger(__name__)
MAX_RETRY = 10
headers = config.headers
@@ -78,6 +78,9 @@ class ReplayChatAsync:
direct_mode = False):
self.video_id = video_id
self.seektime = seektime
if isinstance(processor, tuple):
self.processor = Combinator(processor)
else:
self.processor = processor
self._buffer = buffer
self._callback = callback
@@ -194,13 +197,15 @@ class ReplayChatAsync:
await asyncio.sleep(diff_time)
continuation = metadata.get('continuation')
except ChatParseException as e:
logger.info(f"{str(e)}video_id:\"{self.video_id}\"")
logger.error(f"{str(e)}video_id:\"{self.video_id}\"")
return
except (TypeError , json.JSONDecodeError) :
logger.error(f"{traceback.format_exc(limit = -1)}")
self.terminate()
return
logger.debug(f"[{self.video_id}]チャット取得を終了しました。")
self.terminate()
async def _get_livechat_json(self, continuation, session, headers):
'''
@@ -282,7 +287,7 @@ class ReplayChatAsync:
if self._direct_mode == False:
#bufferにダミーオブジェクトを入れてis_alive()を判定させる
self._buffer.put_nowait({'chatdata':'','timeout':1})
logger.info(f'終了しました:[{self.video_id}]')
logger.info(f'[{self.video_id}]終了しました')
@classmethod
def _set_exception_handler(cls, handler):

View File

@@ -10,15 +10,14 @@ from concurrent.futures import CancelledError, ThreadPoolExecutor
from .buffer import Buffer
from ..parser.live import Parser
from .. import config
from .. import mylogger
from ..exceptions import ChatParseException,IllegalFunctionCall
from ..paramgen import liveparam
from ..processors.default.processor import DefaultProcessor
from ..processors.combinator import Combinator
logger = mylogger.get_logger(__name__,mode=config.LOGGER_MODE)
MAX_RETRY = 10
logger = config.logger(__name__)
headers = config.headers
MAX_RETRY = 10
class LiveChat:
@@ -65,13 +64,16 @@ class LiveChat:
_listeners= []
def __init__(self, video_id,
processor = DefaultProcessor(),
buffer = Buffer(maxsize = 20),
buffer = None,
interruptable = True,
callback = None,
done_callback = None,
direct_mode = False
):
self.video_id = video_id
if isinstance(processor, tuple):
self.processor = Combinator(processor)
else:
self.processor = processor
self._buffer = buffer
self._callback = callback
@@ -142,8 +144,8 @@ class LiveChat:
def _listen(self, continuation):
''' continuationに紐付いたチャットデータを取得し
BUfferにチャットデータを格納、
次のcontinuaitonを取得してループする
Bufferにチャットデータを格納、
次のcontinuaitonを取得してループする
Parameter
---------
@@ -175,9 +177,11 @@ class LiveChat:
time.sleep(diff_time)
continuation = metadata.get('continuation')
except ChatParseException as e:
logger.info(f"{str(e)}video_id:\"{self.video_id}\"")
self.terminate()
logger.error(f"{str(e)}video_id:\"{self.video_id}\"")
return
except (TypeError , json.JSONDecodeError) :
self.terminate()
logger.error(f"{traceback.format_exc(limit = -1)}")
return
@@ -206,6 +210,7 @@ class LiveChat:
else:
logger.error(f"[{self.video_id}]"
f"Exceeded retry count. status_code={status_code}")
self.terminate()
return None
return livechat_json
@@ -254,18 +259,10 @@ class LiveChat:
if self._direct_mode == False:
#bufferにダミーオブジェクトを入れてis_alive()を判定させる
self._buffer.put({'chatdata':'','timeout':1})
logger.info(f'終了しました:[{self.video_id}]')
logger.info(f'[{self.video_id}]終了しました')
@classmethod
def shutdown(cls, event, sig = None, handler=None):
logger.debug("シャットダウンしています")
for t in LiveChat._listeners:
t._is_alive = False

View File

@@ -11,15 +11,14 @@ from queue import Queue
from .buffer import Buffer
from ..parser.replay import Parser
from .. import config
from .. import mylogger
from ..exceptions import ChatParseException,IllegalFunctionCall
from ..paramgen import arcparam
from ..processors.default.processor import DefaultProcessor
from ..processors.combinator import Combinator
logger = mylogger.get_logger(__name__,mode=config.LOGGER_MODE)
MAX_RETRY = 10
logger = config.logger(__name__)
headers = config.headers
MAX_RETRY = 10
class ReplayChat:
@@ -61,7 +60,7 @@ class ReplayChat:
チャットデータ取得ループ_listen用のスレッド
_is_alive : bool
チャット取得を終了したか
チャット取得を停止するためのフラグ
'''
_setup_finished = False
@@ -70,7 +69,7 @@ class ReplayChat:
def __init__(self, video_id,
seektime = 0,
processor = DefaultProcessor(),
buffer = Buffer(maxsize = 20),
buffer = None,
interruptable = True,
callback = None,
done_callback = None,
@@ -78,6 +77,9 @@ class ReplayChat:
):
self.video_id = video_id
self.seektime = seektime
if isinstance(processor, tuple):
self.processor = Combinator(processor)
else:
self.processor = processor
self._buffer = buffer
self._callback = callback
@@ -90,6 +92,7 @@ class ReplayChat:
self._pauser.put_nowait(None)
self._setup()
if not ReplayChat._setup_finished:
ReplayChat._setup_finished = True
if interruptable:
@@ -150,7 +153,7 @@ class ReplayChat:
def _listen(self, continuation):
''' continuationに紐付いたチャットデータを取得し
にチャットデータを格納、
BUfferにチャットデータを格納、
次のcontinuaitonを取得してループする
Parameter
@@ -189,9 +192,11 @@ class ReplayChat:
time.sleep(diff_time)
continuation = metadata.get('continuation')
except ChatParseException as e:
logger.error(f"{str(e)}動画ID:\"{self.video_id}\"")
self.terminate()
logger.error(f"{str(e)}video_id:\"{self.video_id}\"")
return
except (TypeError , json.JSONDecodeError) :
self.terminate()
logger.error(f"{traceback.format_exc(limit = -1)}")
return
@@ -220,6 +225,7 @@ class ReplayChat:
else:
logger.error(f"[{self.video_id}]"
f"Exceeded retry count. status_code={status_code}")
self.terminate()
return None
return livechat_json
@@ -266,7 +272,7 @@ class ReplayChat:
'''Listener終了時のコールバック'''
try:
self.terminate()
except CancelledError:
except RuntimeError:
logger.debug(f'[{self.video_id}]cancelled:{sender}')
def terminate(self):
@@ -277,7 +283,7 @@ class ReplayChat:
if self._direct_mode == False:
#bufferにダミーオブジェクトを入れてis_alive()を判定させる
self._buffer.put({'chatdata':'','timeout':1})
logger.info(f'終了しました:[{self.video_id}]')
logger.info(f'[{self.video_id}]終了しました')
@classmethod
def shutdown(cls, event, sig = None, handler=None):

View File

@@ -6,14 +6,13 @@ This module is parser of live chat JSON.
import json
from .. import config
from .. import mylogger
from .. exceptions import (
ResponseContextError,
NoContentsException,
NoContinuationsException )
logger = mylogger.get_logger(__name__,mode=config.LOGGER_MODE)
logger = config.logger(__name__)
class Parser:
@@ -59,7 +58,7 @@ class Parser:
if metadata is None:
unknown = list(cont.keys())[0]
if unknown:
logger.error(f"Received unknown continuation type:{unknown}")
logger.debug(f"Received unknown continuation type:{unknown}")
metadata = cont.get(unknown)
metadata.setdefault('timeoutMs', 10000)
chatdata = contents['liveChatContinuation'].get('actions')

View File

@@ -1,14 +1,12 @@
import json
from .. import config
from .. import mylogger
from .. exceptions import (
ResponseContextError,
NoContentsException,
NoContinuationsException )
logger = mylogger.get_logger(__name__,mode=config.LOGGER_MODE)
logger = config.logger(__name__)
class Parser:
def parse(self, jsn):
@@ -53,12 +51,13 @@ class Parser:
metadata = cont.get('liveChatReplayContinuationData')
if metadata is None:
unknown = list(cont.keys())[0]
if unknown:
if unknown != "playerSeekContinuationData":
logger.debug(f"Received unknown continuation type:{unknown}")
metadata = cont.get(unknown)
actions = contents['liveChatContinuation'].get('actions')
if actions is None:
raise NoContentsException('チャットデータを取得できませんでした。')
#後続のチャットデータなし
return {"continuation":None,"timeout":0,"chatdata":[]}
interval = self.get_interval(actions)
metadata.setdefault("timeoutMs",interval)
"""アーカイブ済みチャットはライブチャットと構造が異なっているため、以下の行により

View File

@@ -10,14 +10,14 @@ class ChatProcessor:
Parameter
----------
chat_components: [LIST:component]
chat_components: List[component]
component : dict {
"video_id" : str
動画ID
"timeout" : int
次のチャットの再読み込みまでの時間(秒)
"chatdata" : list<object>
チャットデータactionsのリスト
"chatdata" : List[dict]
チャットデータのリスト
}
'''
pass

View File

@@ -0,0 +1,39 @@
from .chat_processor import ChatProcessor
class Combinator(ChatProcessor):
'''
Combinator combines multiple chat processors.
Specify processors as tuple at `processor` params of LiveChat object.
For example:
[constructor]
chat = LiveChat("video_id", processor = ( Processor1(), Processor2(), Processor3() )
[receive return values]
ret1, ret2, ret3 = chat.get()
The return values are tuple of processed chat data,
the order of return depends on parameter order.
Parameter
---------
processors : Tuple[ChatProcessor]
multiple processors for processing chat data
'''
def __init__(self, processors: tuple):
self.processors = processors
def process(self, chat_components: list):
'''
Called from LiveChat.get() function by user,
or LiveChat._listen() automatically.
Returns
-------
Tuple of chat data processed by each chat processor.
'''
return tuple([processor.process(chat_components)
for processor in self.processors])

View File

@@ -4,11 +4,11 @@ from .renderer.textmessage import LiveChatTextMessageRenderer
from .renderer.paidmessage import LiveChatPaidMessageRenderer
from .renderer.paidsticker import LiveChatPaidStickerRenderer
from .renderer.legacypaid import LiveChatLegacyPaidMessageRenderer
from ... import mylogger
from .. chat_processor import ChatProcessor
from ... import config
logger = mylogger.get_logger(__name__,mode=config.LOGGER_MODE)
logger = config.logger(__name__)
class CompatibleProcessor:
class CompatibleProcessor(ChatProcessor):
def process(self, chat_components: list):

View File

@@ -4,9 +4,9 @@ from .renderer.textmessage import LiveChatTextMessageRenderer
from .renderer.paidmessage import LiveChatPaidMessageRenderer
from .renderer.paidsticker import LiveChatPaidStickerRenderer
from .renderer.legacypaid import LiveChatLegacyPaidMessageRenderer
from .. chat_processor import ChatProcessor
from ... import config
from ... import mylogger
logger = mylogger.get_logger(__name__,mode=config.LOGGER_MODE)
logger = config.logger(__name__)
class Chatdata:
def __init__(self,chatlist:list, timeout:float):
@@ -25,7 +25,7 @@ class Chatdata:
return
await asyncio.sleep(self.interval/len(self.items))
class DefaultProcessor:
class DefaultProcessor(ChatProcessor):
def process(self, chat_components: list):
chatlist = []

View File

@@ -19,8 +19,7 @@ class BaseRenderer:
else:
self.elapsedTime = ""
self.datetime = self.get_datetime(timestampUsec)
self.message = self.get_message(self.renderer)
self.messageEx = self.get_message_ex(self.renderer)
self.message ,self.messageEx = self.get_message(self.renderer)
self.id = self.renderer.get('id')
self.amountValue= 0.0
self.amountString = ""
@@ -44,6 +43,7 @@ class BaseRenderer:
def get_message(self,renderer):
message = ''
message_ex = []
if renderer.get("message"):
runs=renderer["message"].get("runs")
if runs:
@@ -51,22 +51,13 @@ class BaseRenderer:
if r:
if r.get('emoji'):
message += r['emoji'].get('shortcuts',[''])[0]
message_ex.append(r['emoji']['image']['thumbnails'][1].get('url'))
else:
message += r.get('text','')
return message
message_ex.append(r.get('text',''))
return message, message_ex
def get_message_ex(self,renderer):
message = []
if renderer.get("message"):
runs=renderer["message"].get("runs")
if runs:
for r in runs:
if r:
if r.get('emoji'):
message.append(r['emoji']['image']['thumbnails'][1].get('url'))
else:
message.append(r.get('text',''))
return message
def get_badges(self,renderer):
isVerified = False

View File

@@ -0,0 +1,8 @@
from .chat_processor import ChatProcessor
class DummyProcessor(ChatProcessor):
'''
Dummy processor just returns received chat_components directly.
'''
def process(self, chat_components: list):
return chat_components

View File

@@ -4,7 +4,7 @@ speedmeter.py
Calculate speed of chat.
"""
import calendar, datetime, pytz
from .chat_processor import ChatProcessor
class RingQueue:
"""
リング型キュー
@@ -21,7 +21,7 @@ class RingQueue:
キュー内に余裕があるか。キュー内のアイテム個数が、キューの最大個数未満であればTrue。
"""
def __init__(self, capacity = 10):
def __init__(self, capacity):
"""
コンストラクタ
@@ -77,34 +77,39 @@ class RingQueue:
def item_count(self):
return len(self.items)
class SpeedCalculator(RingQueue):
class SpeedCalculator(ChatProcessor, RingQueue):
"""
チャットの勢いを計算するクラス
チャットの勢いを計算する
一定期間のチャットデータのうち、最初のチャットの投稿時刻と
最後のチャットの投稿時刻の差を、チャット数で割り返し
1分あたりの速度に換算する。
Parameter
----------
格納するチャットブロックの数
capacity : int
RingQueueに格納するチャット勢い算出用データの最大数
"""
def __init__(self, capacity, video_id):
def __init__(self, capacity = 10):
super().__init__(capacity)
self.video_id=video_id
self.speed = 0
def process(self, chat_components: list):
chatdata = []
if chat_components:
for component in chat_components:
if component.get("chatdata"):
chatdata.extend(component.get("chatdata"))
chatdata = component.get('chatdata')
if chatdata is None:
return self.speed
self.speed = self.calc(chatdata)
self._put_chatdata(chatdata)
self.speed = self._calc_speed()
return self.speed
def _value(self):
def _calc_speed(self):
"""
ActionsQueue内のチャットデータリストから
RingQueue内のチャット勢い算出用データリストを元に
チャット速度を計算して返す
Return
@@ -112,7 +117,7 @@ class SpeedCalculator(RingQueue):
チャット速度(1分間で換算したチャット数)
"""
try:
#キュー内のactionsの総チャット数
#キュー内の総チャット数
total = sum(item['chat_count'] for item in self.items)
#キュー内の最初と最後のチャットの時間差
duration = (self.items[self.last_pos]['endtime']
@@ -123,24 +128,20 @@ class SpeedCalculator(RingQueue):
except IndexError:
return 0
def _get_timestamp(self, action :dict):
def _put_chatdata(self, actions):
"""
チャットデータのtimestampUsecを読み取る
liveChatTickerSponsorItemRenderer等のtickerデータは時刻格納位置が
異なるため、時刻データなしとして扱う
チャットデータからタイムスタンプを読み取り、勢い測定用のデータを組み立て、
RingQueueに投入する。
200円以上のスパチャはtickerとmessageの2つのデータが生成されるが、
tickerの方は時刻データの場所が異なることを利用し、勢いの集計から除外している。
Parameter
---------
actions : List[dict]
チャットデータ(addChatItemAction) のリスト
"""
try:
item = action['addChatItemAction']['item']
timestamp = int(item[list(item.keys())[0]]['timestampUsec'])
except (KeyError,TypeError):
return None
return timestamp
def calc(self,actions):
def empty_data():
def _put_emptydata():
'''
データがない場合にゼロのデータをリングキューに入れる
チャットデータがない場合にのデータをキューに投入する。
'''
timestamp_now = calendar.timegm(datetime.datetime.
now(pytz.utc).utctimetuple())
@@ -149,12 +150,23 @@ class SpeedCalculator(RingQueue):
'starttime':int(timestamp_now),
'endtime':int(timestamp_now)
})
return self._value()
def _get_timestamp(action :dict):
"""
チャットデータから時刻データを取り出す。
"""
try:
item = action['addChatItemAction']['item']
timestamp = int(item[list(item.keys())[0]]['timestampUsec'])
except (KeyError,TypeError):
return None
return timestamp
if actions is None or len(actions)==0:
return empty_data
_put_emptydata()
return
#actions内の時刻データを持つチャットデータの数tickerは除く
#actions内の時刻データを持つチャットデータの数
counter=0
#actions内の最初のチャットデータの時刻
starttime= None
@@ -163,7 +175,7 @@ class SpeedCalculator(RingQueue):
for action in actions:
#チャットデータからtimestampUsecを読み取る
gettime = self._get_timestamp(action)
gettime = _get_timestamp(action)
#時刻のないデータだった場合は次の行のデータで読み取り試行
if gettime is None:
@@ -179,9 +191,10 @@ class SpeedCalculator(RingQueue):
#チャットの数をインクリメント
counter += 1
#チャット速度用のデータをリングキューに送る
#チャット速度用のデータをRingQueueに送る
if starttime is None or endtime is None:
return empty_data
_put_emptydata()
return
self.put({
'chat_count':counter,
@@ -189,4 +202,3 @@ class SpeedCalculator(RingQueue):
'endtime':int(endtime/1000000)
})
return self._value()

View File

@@ -9,7 +9,7 @@ def download(url):
json.dump(html.json(),f,ensure_ascii=False)
def save(data,filename):
with open(str(datetime.datetime.now().strftime('%Y-%m-%d %H-%M-%S')
)+filename,mode ='w',encoding='utf-8') as f:
def save(data,filename,extention):
with open(filename+"_"+(datetime.datetime.now().strftime('%Y-%m-%d %H-%M-%S')
)+extention,mode ='w',encoding='utf-8') as f:
f.writelines(data)

View File

@@ -1,6 +1,6 @@
from setuptools import setup, find_packages, Command
from codecs import open
from os import path, system
#from codecs import open as open_c
from os import path, system, remove, rename
import re
package_name = "pytchat"
@@ -28,6 +28,16 @@ assert author
assert author_email
assert url
with open('README.MD', 'r', encoding='utf-8') as f:
txt = f.read()
with open('README1.MD', 'w', encoding='utf-8', newline='\n') as f:
f.write(txt)
remove("README.MD")
rename("README1.MD","README.MD")
with open('README.md', encoding='utf-8') as f:
long_description = f.read()

View File

@@ -0,0 +1,68 @@
import json
import pytest
import asyncio,aiohttp
from pytchat.parser.live import Parser
from pytchat.processors.compatible.processor import CompatibleProcessor
from pytchat.exceptions import (
NoLivechatRendererException,NoYtinitialdataException,
ResponseContextError, NoContentsException)
from pytchat.processors.speed_calculator import SpeedCalculator
parser = Parser()
def test_speed_1(mocker):
'''test speed calculation with normal json.
test json has 15 chatdata, duration is 30 seconds,
so the speed of chatdata is 30 chats/minute.
'''
processor = SpeedCalculator(capacity=30)
_json = _open_file("tests/testdata/speed/speedtest_normal.json")
_, chatdata = parser.parse(json.loads(_json))
data = {
"video_id" : "",
"timeout" : 10,
"chatdata" : chatdata
}
ret = processor.process([data])
assert 30 == ret
def test_speed_2(mocker):
'''test speed calculation with no valid chat data.
'''
processor = SpeedCalculator(capacity=30)
_json = _open_file("tests/testdata/speed/speedtest_undefined.json")
_, chatdata = parser.parse(json.loads(_json))
data = {
"video_id" : "",
"timeout" : 10,
"chatdata" : chatdata
}
ret = processor.process([data])
assert 0 == ret
def test_speed_3(mocker):
'''test speed calculation with empty data.
'''
processor = SpeedCalculator(capacity=30)
_json = _open_file("tests/testdata/speed/speedtest_empty.json")
_, chatdata = parser.parse(json.loads(_json))
data = {
"video_id" : "",
"timeout" : 10,
"chatdata" : chatdata
}
ret = processor.process([data])
assert 0 == ret
def _open_file(path):
with open(path,mode ='r',encoding = 'utf-8') as f:
return f.read()

164
tests/testdata/chat.json vendored Normal file
View File

@@ -0,0 +1,164 @@
{
"timing": {
"info": {
"st": 164
}
},
"csn": "",
"response": {
"responseContext": {
"serviceTrackingParams": [{
"service": "CSI",
"params": [{
"key": "GetLiveChat_rid",
"value": ""
}, {
"key": "c",
"value": "WEB"
}, {
"key": "cver",
"value": "2.20191219.03.01"
}, {
"key": "yt_li",
"value": "0"
}]
}, {
"service": "GFEEDBACK",
"params": [{
"key": "e",
"value": ""
}, {
"key": "logged_in",
"value": "0"
}]
}, {
"service": "GUIDED_HELP",
"params": [{
"key": "logged_in",
"value": "0"
}]
}, {
"service": "ECATCHER",
"params": [{
"key": "client.name",
"value": "WEB"
}, {
"key": "client.version",
"value": "2.2"
}, {
"key": "innertube.build.changelist",
"value": "228"
}, {
"key": "innertube.build.experiments.source_version",
"value": "2858"
}, {
"key": "innertube.build.label",
"value": "youtube.ytfe.innertube_"
}, {
"key": "innertube.build.timestamp",
"value": "154"
}, {
"key": "innertube.build.variants.checksum",
"value": "e"
}, {
"key": "innertube.run.job",
"value": "ytfe-innertube-replica-only.ytfe"
}]
}],
"webResponseContextExtensionData": {
"ytConfigData": {
"csn": "ADw",
"visitorData": "%3D%3D"
}
}
},
"continuationContents": {
"liveChatContinuation": {
"continuations": [{
"timedContinuationData": {
"timeoutMs": 10000,
"continuation": "continuation"
}
}],
"actions": [{
"addChatItemAction": {
"item": {
"liveChatTextMessageRenderer": {
"message": {
"runs": [{
"text": "message"
}]
},
"authorName": {
"simpleText": "authorName"
},
"authorPhoto": {
"thumbnails": [{
"url": "https://yt3.ggpht.com/photo.jpg",
"width": 32,
"height": 32
}, {
"url": "https://yt3.ggpht.com/photo.jpg",
"width": 64,
"height": 64
}]
},
"contextMenuEndpoint": {
"commandMetadata": {
"webCommandMetadata": {
"ignoreNavigation": true
}
},
"liveChatItemContextMenuEndpoint": {
"params": "params"
}
},
"id": "id",
"timestampUsec": "1576851922945411",
"authorBadges": [{
"liveChatAuthorBadgeRenderer": {
"customThumbnail": {
"thumbnails": [{
"url": "https://yt3.ggpht.com/photo.jpg"
}, {
"url": "https://yt3.ggpht.com/photo.jpg"
}]
},
"tooltip": "メンバー6 か月)",
"accessibility": {
"accessibilityData": {
"label": "メンバー6 か月)"
}
}
}
}],
"authorExternalChannelId": "UC",
"contextMenuAccessibility": {
"accessibilityData": {
"label": "コメントの操作"
}
}
}
},
"clientId": "00000000000000000000"
}
}
]}
},
"xsrf_token": "xsrf_token",
"url": "/live_chat/get_live_chat?continuation=0",
"endpoint": {
"commandMetadata": {
"webCommandMetadata": {
"url": "/live_chat/get_live_chat?continuation=0",
"rootVe": 0
}
},
"urlEndpoint": {
"url": "/live_chat/get_live_chat?continuation=0"
}
}
}
}

View File

@@ -0,0 +1,24 @@
{
"timing": {
"info": {
"st": 164
}
},
"csn": "",
"response": {
"responseContext": {
},
"continuationContents": {
"liveChatContinuation": {
"continuations": [{
"timedContinuationData": {
"timeoutMs": 10000,
"continuation": "continuation"
}
}]
}
}
}
}

View File

@@ -0,0 +1,188 @@
{
"timing": {
"info": {
"st": 164
}
},
"csn": "",
"response": {
"responseContext": {
},
"continuationContents": {
"liveChatContinuation": {
"continuations": [{
"timedContinuationData": {
"timeoutMs": 10000,
"continuation": "continuation"
}
}],
"actions": [{
"addChatItemAction": {
"item": {
"liveChatTextMessageRenderer": {
"timestampUsec": "1500000000000000"
}
}
}
},
{
"addChatItemAction": {
"item": {
"liveChatTextMessageRenderer": {
"timestampUsec": "1500000000000000"
}
}
}
},
{
"addChatItemAction": {
"item": {
"liveChatTextMessageRenderer": {
"timestampUsec": "1500000000000000"
}
}
}
},
{
"addChatItemAction": {
"item": {
"liveChatTextMessageRenderer": {
"timestampUsec": "1500000000000000"
}
}
}
},
{
"addChatItemAction": {
"item": {
"liveChatTextMessageRenderer": {
"timestampUsec": "1500000000000000"
}
}
}
},
{
"addChatItemAction": {
"item": {
"liveChatTextMessageRenderer": {
"timestampUsec": "1500000000000000"
}
}
}
},
{
"addChatItemAction": {
"item": {
"liveChatTextMessageRenderer": {
"timestampUsec": "1500000000000000"
}
}
}
},
{
"addChatItemAction": {
"item": {
"liveChatTextMessageRenderer": {
"timestampUsec": "1500000000000000"
}
}
}
},
{
"addChatItemAction": {
"item": {
"liveChatTextMessageRenderer": {
"timestampUsec": "1500000000000000"
}
}
}
},
{
"addChatItemAction": {
"item": {
"liveChatTextMessageRenderer": {
"timestampUsec": "1500000000000000"
}
}
}
},
{
"addChatItemAction": {
"item": {
"liveChatTextMessageRenderer": {
"timestampUsec": "1500000000000000"
}
}
}
},
{
"addChatItemAction": {
"item": {
"liveChatTextMessageRenderer": {
"timestampUsec": "1500000000000000"
}
}
}
},
{
"addChatItemAction": {
"item": {
"liveChatTextMessageRenderer": {
"timestampUsec": "1500000000000000"
}
}
}
},
{
"addChatItemAction": {
"item": {
"liveChatTextMessageRenderer": {
"timestampUsec": "1500000000000000"
}
}
}
},
{
"addChatItemAction": {
"item": {
"liveChatTextMessageRenderer": {
"timestampUsec": "1500000030000000"
}
}
}
}
]}
}
}
}

View File

@@ -0,0 +1,42 @@
{
"timing": {
"info": {
"st": 164
}
},
"csn": "",
"response": {
"responseContext": {
},
"continuationContents": {
"liveChatContinuation": {
"continuations": [{
"timedContinuationData": {
"timeoutMs": 10000,
"continuation": "continuation"
}
}],
"actions": [{
"addChatItemAction": {
"liveChatPlaceholderItemRenderer": {
"id": "",
"timestampUsec": "1500000000000000"
}
}
},
{
"addChatItemAction": {
"item": {
"liveChatPlaceholderItemRenderer": {
"id": "",
"timestampUsec": "1500000030000000"
}
}
}
}
]}
}
}
}