Merge branch 'hotfix/fix_json'

pytchat/__init__.py

@@ -2,7 +2,7 @@
 pytchat is a lightweight python library to browse youtube livechat without Selenium or BeautifulSoup.
 """
 __copyright__    = 'Copyright (C) 2019 taizan-hokuto'
-__version__      = '0.2.2'
+__version__      = '0.3.2'
 __license__      = 'MIT'
 __author__       = 'taizan-hokuto'
 __author_email__ = '55448286+taizan-hokuto@users.noreply.github.com'

pytchat/cli/__init__.py

@@ -1,13 +1,17 @@
 import argparse
-
+import asyncio
+try:
+    from asyncio import CancelledError
+except ImportError:
+    from asyncio.futures import CancelledError
 import os
 import signal
-import time
 from json.decoder import JSONDecodeError
 from pathlib import Path
+from httpcore import ReadTimeout as HCReadTimeout, NetworkError as HCNetworkError
 from .arguments import Arguments
 from .progressbar import ProgressBar
-from .. exceptions import InvalidVideoIdException, NoContents, PatternUnmatchError
+from .. exceptions import InvalidVideoIdException, NoContents, PatternUnmatchError, UnknownConnectionError
 from .. processors.html_archiver import HTMLArchiver
 from .. tool.extract.extractor import Extractor
 from .. tool.videoinfo import VideoInfo
@@ -37,6 +41,7 @@ def main():
                         help='Save error data when error occurs(".dat" file)')
     parser.add_argument(f'--{Arguments.Name.VERSION}', action='store_true',
                         help='Show version')
+
     Arguments(parser.parse_args().__dict__)
 
     if Arguments().print_version:
@@ -47,75 +52,106 @@ def main():
     if not Arguments().video_ids:
         parser.print_help()
         return
-    for counter, video_id in enumerate(Arguments().video_ids):
-        if '[' in video_id:
-            video_id = video_id.replace('[', '').replace(']', '')
-        try:
-            video_id = extract_video_id(video_id)
-            if os.path.exists(Arguments().output):
-                path = Path(Arguments().output + video_id + '.html')
-            else:
-                raise FileNotFoundError
-            err = None
-            for _ in range(3): # retry 3 times
-                try:                
-                    info = VideoInfo(video_id)
-                    break
-                except (PatternUnmatchError, JSONDecodeError, InvalidVideoIdException) as e:
-                    err = e
-                    time.sleep(2)
-                    continue
-            else:
-                print("Cannot parse video information.:{}".format(video_id))
-                if Arguments().save_error_data:
-                    util.save(err.doc, "ERR", ".dat")
-                continue
 
+    if not os.path.exists(Arguments().output):
+        print("\nThe specified directory does not exist.:{}\n".format(Arguments().output))
+        return
+    try:
+        Runner().run()
+    except CancelledError as e:
+        print(str(e))
+
+
+class Runner:
+    
+    def run(self) -> None:
+        ex = None
+        pbar = None
+        for counter, video_id in enumerate(Arguments().video_ids):
             if len(Arguments().video_ids) > 1:
                 print(f"\n{'-' * 10} video:{counter + 1} of {len(Arguments().video_ids)} {'-' * 10}")
-            print(f"\n"
-                  f" video_id: {video_id}\n"
-                  f" channel:  {info.get_channel_name()}\n"
-                  f" title:    {info.get_title()}")
 
-            print(f" output path: {path.resolve()}")
+            try:
-            duration = info.get_duration()
+                video_id = extract_video_id(video_id)
-            pbar = ProgressBar(total=(duration * 1000), status="Extracting")
+                separated_path = str(Path(Arguments().output)) + os.path.sep
-            ex = Extractor(video_id,               
+                path = util.checkpath(separated_path + video_id + '.html')
-                    callback=pbar._disp,
+                try:
-                    div=10)
+                    info = VideoInfo(video_id)
-            signal.signal(signal.SIGINT, (lambda a, b: cancel(ex, pbar)))
+                except Exception as e:
-            data = ex.extract()
+                    print("Cannot parse video information.:{} {}".format(video_id, type(e)))
-            if data == []:
+                    if Arguments().save_error_data:
-                return False
+                        util.save(str(e), "ERR", ".dat")
-            pbar.reset("#", "=", total=len(data), status="Rendering  ")
+                    continue
-            processor = HTMLArchiver(Arguments().output + video_id + '.html', callback=pbar._disp)
-            processor.process(
-                [{'video_id': None,
-                'timeout': 1,
-                'chatdata': (action["replayChatItemAction"]["actions"][0] for action in data)}]
-            )
-            processor.finalize()
-            pbar.reset('#', '#', status='Completed   ')
-            pbar.close()
-            print()
-            if pbar.is_cancelled():
-                print("\nThe extraction process has been discontinued.\n")
-        except InvalidVideoIdException:
-            print("Invalid Video ID or URL:", video_id)
-        except NoContents as e:
-            print(e)
-        except FileNotFoundError:
-            print("The specified directory does not exist.:{}".format(Arguments().output))
-        except JSONDecodeError as e:
-            print(e.msg)
-            print("JSONDecodeError.:{}".format(video_id))
-            if Arguments().save_error_data:
-                util.save(e.doc, "ERR_JSON_DECODE", ".dat")
 
-    return
+                print(f"\n"
+                    f" video_id: {video_id}\n"
+                    f" channel:  {info.get_channel_name()}\n"
+                    f" title:    {info.get_title()}\n"
+                    f" output path: {path}")
+
+                duration = info.get_duration()
+                pbar = ProgressBar(total=(duration * 1000), status_txt="Extracting")
+                ex = Extractor(video_id,
+                        callback=pbar.disp,
+                        div=10)
+                signal.signal(signal.SIGINT, (lambda a, b: self.cancel(ex, pbar)))
+
+                data = ex.extract()
+                if data == []:
+                    continue
+                pbar.reset("#", "=", total=len(data), status_txt="Rendering  ")
+                processor = HTMLArchiver(path, callback=pbar.disp)
+                processor.process(
+                    [{'video_id': None,
+                    'timeout': 1,
+                    'chatdata': (action["replayChatItemAction"]["actions"][0] for action in data)}]
+                )
+                processor.finalize()
+                pbar.reset('#', '#', status_txt='Completed   ')
+                pbar.close()
+                print()
+                if pbar.is_cancelled():
+                    print("\nThe extraction process has been discontinued.\n")
+            except InvalidVideoIdException:
+                print("Invalid Video ID or URL:", video_id)
+            except NoContents as e:
+                print(f"Abort:{str(e)}:[{video_id}]")
+            except (JSONDecodeError, PatternUnmatchError) as e:
+                print("{}:{}".format(e.msg, video_id))
+                if Arguments().save_error_data:
+                    util.save(e.doc, "ERR_", ".dat")
+            except (UnknownConnectionError, HCNetworkError, HCReadTimeout) as e:
+                print(f"An unknown network error occurred during the processing of [{video_id}]. : " + str(e))
+            except Exception as e:
+                print(f"Abort:{str(type(e))} {str(e)[:80]}")
+            finally:
+                clear_tasks()
+
+        return
+
+    def cancel(self, ex=None, pbar=None) -> None:
+        '''Called when keyboard interrupted has occurred.
+        '''
+        print("\nKeyboard interrupted.\n")
+        if ex and pbar:
+            ex.cancel()
+            pbar.cancel()
 
 
-def cancel(ex, pbar):
+def clear_tasks():
-    ex.cancel()
+    '''
-    pbar.cancel()
+    Clear remained tasks.
+    Called when internal exception has occurred or
+    after each extraction process is completed.
+    '''
+    async def _shutdown():
+        tasks = [t for t in asyncio.all_tasks()
+                if t is not asyncio.current_task()]
+        for task in tasks:
+            task.cancel()
+            
+    try:
+        loop = asyncio.get_event_loop()
+        loop.run_until_complete(_shutdown())
+    except Exception as e:
+        print(e)

pytchat/cli/progressbar.py

@@ -9,21 +9,20 @@ import sys
 
 
 class ProgressBar:
-    def __init__(self, total, status):
+    def __init__(self, total, status_txt):
         self._bar_len = 60
         self._cancelled = False
-        self.reset(total=total, status=status)
+        self.reset(total=total, status_txt=status_txt)
-        self._blinker = 0
         
-    def reset(self, symbol_done="=", symbol_space=" ", total=100, status=''):
+    def reset(self, symbol_done="=", symbol_space=" ", total=100, status_txt=''):
-        self.con_width = shutil.get_terminal_size(fallback=(80, 24)).columns
+        self._console_width = shutil.get_terminal_size(fallback=(80, 24)).columns
         self._symbol_done = symbol_done
         self._symbol_space = symbol_space
         self._total = total
-        self._status = status
+        self._status_txt = status_txt
         self._count = 0
 
-    def _disp(self, _, fetched):
+    def disp(self, _, fetched):
         self._progress(fetched, self._total)
 
     def _progress(self, fillin, total):
@@ -39,11 +38,10 @@ class ProgressBar:
             
         bar = self._symbol_done * filled_len + \
               self._symbol_space * (self._bar_len - filled_len)
-        disp = f" [{bar}] {percents:>5.1f}% ...{self._status} "[:self.con_width - 1] + '\r'
+        disp = f" [{bar}] {percents:>5.1f}% ...{self._status_txt} "[:self._console_width - 1] + '\r'
 
         sys.stdout.write(disp)
         sys.stdout.flush()
-        self._blinker += 1
 
     def close(self):
         if not self._cancelled:

pytchat/exceptions.py

@@ -43,7 +43,6 @@ class InvalidVideoIdException(Exception):
         self.doc = doc
 
 
-
 class UnknownConnectionError(Exception):
     pass
 

pytchat/processors/html_archiver.py

@@ -1,9 +1,12 @@
+import httpx
 import os
 import re
-import httpx
+import time
 from base64 import standard_b64encode
+from httpx import NetworkError, ReadTimeout
 from .chat_processor import ChatProcessor
 from .default.processor import DefaultProcessor
+from ..exceptions import UnknownConnectionError
 
 
 PATTERN = re.compile(r"(.*)\(([0-9]+)\)$")
@@ -112,7 +115,18 @@ class HTMLArchiver(ChatProcessor):
                        for item in message_items)
 
     def _encode_img(self, url):
-        resp = httpx.get(url)
+        err = None
+        for _ in range(5):
+            try:
+                resp = httpx.get(url, timeout=30)
+                break
+            except (NetworkError, ReadTimeout) as e:
+                print("Network Error. retrying...")
+                err = e
+                time.sleep(3)
+        else:
+            raise UnknownConnectionError(str(err))
+
         return standard_b64encode(resp.content).decode()
 
     def _set_emoji_table(self, item: dict):

pytchat/tool/extract/asyncdl.py

@@ -1,5 +1,6 @@
-import httpx
 import asyncio
+import httpx
+import socket
 from . import parser
 from . block import Block
 from . worker import ExtractWorker
@@ -8,14 +9,19 @@ from ... import config
 from ... paramgen import arcparam
 from ... exceptions import UnknownConnectionError
 from concurrent.futures import CancelledError
+from httpx import NetworkError, TimeoutException, ConnectError
 from json import JSONDecodeError
 from urllib.parse import quote
 
+
 headers = config.headers
 REPLAY_URL = "https://www.youtube.com/live_chat_replay/" \
              "get_live_chat_replay?continuation="
 MAX_RETRY_COUNT = 3
 
+# Set to avoid duplicate parameters
+param_set = set()
+
 
 def _split(start, end, count, min_interval_sec=120):
     """
@@ -50,6 +56,7 @@ def _split(start, end, count, min_interval_sec=120):
 
 
 def ready_blocks(video_id, duration, div, callback):
+    param_set.clear()
     if div <= 0:
         raise ValueError
 
@@ -62,16 +69,24 @@ def ready_blocks(video_id, duration, div, callback):
     async def _create_block(session, video_id, seektime, callback):
         continuation = arcparam.getparam(video_id, seektime=seektime)
         url = f"{REPLAY_URL}{quote(continuation)}&pbj=1"
+        err = None
         for _ in range(MAX_RETRY_COUNT):
             try:
-                resp = await session.get(url, headers=headers)
+                if continuation in param_set:
+                    next_continuation, actions = None, []
+                    break
+                param_set.add(continuation)
+                resp = await session.get(url, headers=headers, timeout=10)
                 next_continuation, actions = parser.parse(resp.json())
                 break
             except JSONDecodeError:
                 await asyncio.sleep(3)
+            except (NetworkError, TimeoutException, ConnectError) as e:
+                err = e
+                await asyncio.sleep(3)
         else:
             cancel()
-            raise UnknownConnectionError("Abort: Unknown connection error.")
+            raise UnknownConnectionError("Abort:" + str(err))
 
         if actions:
             first = parser.get_offset(actions[0])
@@ -110,16 +125,27 @@ def fetch_patch(callback, blocks, video_id):
 
     async def _fetch(continuation, session) -> Patch:
         url = f"{REPLAY_URL}{quote(continuation)}&pbj=1"
+        err = None
         for _ in range(MAX_RETRY_COUNT):
             try:
+                if continuation in param_set:
+                    continuation, actions = None, []
+                    break
+                param_set.add(continuation)
                 resp = await session.get(url, headers=config.headers)
                 continuation, actions = parser.parse(resp.json())
                 break
             except JSONDecodeError:
                 await asyncio.sleep(3)
+            except (NetworkError, TimeoutException, ConnectError) as e:
+                err = e
+                await asyncio.sleep(3)
+            except socket.error as error:
+                print("socket error", error.errno)
+                await asyncio.sleep(3)
         else:
             cancel()
-            raise UnknownConnectionError("Abort: Unknown connection error.")
+            raise UnknownConnectionError("Abort:" + str(err))
 
         if actions:
             last = parser.get_offset(actions[-1])
@@ -140,15 +166,10 @@ def fetch_patch(callback, blocks, video_id):
 
 
 async def _shutdown():
-    print("\nshutdown...")
     tasks = [t for t in asyncio.all_tasks()
              if t is not asyncio.current_task()]
     for task in tasks:
         task.cancel()
-        try:
-            await task
-        except asyncio.CancelledError:
-            pass
 
 
 def cancel():

pytchat/tool/extract/worker.py

@@ -7,7 +7,6 @@ from typing import Tuple
 class ExtractWorker:
     """
     ExtractWorker associates a download session with a block.
-
     When the worker finishes fetching, the block
     being fetched is splitted and assigned the free worker.
 

pytchat/tool/videoinfo.py

@@ -1,13 +1,16 @@
+import httpx
 import json
 import re
-import httpx
+import time
+from httpx import ConnectError, NetworkError, TimeoutException
 from .. import config
-from ..exceptions import InvalidVideoIdException, PatternUnmatchError
+from ..exceptions import InvalidVideoIdException, PatternUnmatchError, UnknownConnectionError
 from ..util.extract_video_id import extract_video_id
 
-headers = config.headers
 
-pattern = re.compile(r"'PLAYER_CONFIG': ({.*}}})")
+headers = config.headers
+                         
+pattern = re.compile(r"['\"]PLAYER_CONFIG['\"]:\s*({.*})")
 
 item_channel_id = [
     "videoDetails",
@@ -80,19 +83,42 @@ class VideoInfo:
 
     def __init__(self, video_id):
         self.video_id = extract_video_id(video_id)
-        text = self._get_page_text(self.video_id)
+        err = None
-        self._parse(text)
+        for _ in range(3):
+            try:
+                text = self._get_page_text(self.video_id)
+                self._parse(text)
+                break
+            except (InvalidVideoIdException, UnknownConnectionError) as e:
+                print(str(e))
+                raise e
+            except Exception as e:
+                err = e
+                time.sleep(2)
+                pass
+        else:
+            raise err
 
     def _get_page_text(self, video_id):
         url = f"https://www.youtube.com/embed/{video_id}"
-        resp = httpx.get(url, headers=headers)
+        err = None
-        resp.raise_for_status()
+        for _ in range(3):
+            try:
+                resp = httpx.get(url, headers=headers)
+                resp.raise_for_status()
+                break
+            except (ConnectError, NetworkError, TimeoutException) as e:
+                err = e
+                time.sleep(3)
+        else:
+            raise UnknownConnectionError(str(err))
+
         return resp.text
 
     def _parse(self, text):
         result = re.search(pattern, text)
         if result is None:
-            raise PatternUnmatchError(text)
+            raise PatternUnmatchError(doc=text)
         decoder = json.JSONDecoder()
         res = decoder.raw_decode(result.group(1)[:-1])[0]
         response = self._get_item(res, item_response)

pytchat/util/__init__.py

@@ -1,8 +1,12 @@
+import datetime
 import httpx
 import json
-import datetime
+import os
+import re
 from .. import config
 
+PATTERN = re.compile(r"(.*)\(([0-9]+)\)$")
+
 
 def extract(url):
     _session = httpx.Client(http2=True)
@@ -16,3 +20,21 @@ 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)
+
+
+def checkpath(filepath):
+    splitter = os.path.splitext(os.path.basename(filepath))
+    body = splitter[0]
+    extention = splitter[1]
+    newpath = filepath
+    counter = 1
+    while os.path.exists(newpath):
+        match = re.search(PATTERN, body)
+        if match:
+            counter = int(match[2]) + 1
+            num_with_bracket = f'({str(counter)})'
+            body = f'{match[1]}{num_with_bracket}'
+        else:
+            body = f'{body}({str(counter)})'
+        newpath = os.path.join(os.path.dirname(filepath), body + extention)
+    return newpath

pytchat/util/extract_video_id.py

@@ -8,6 +8,9 @@ YT_VIDEO_ID_LENGTH = 11
 
 def extract_video_id(url_or_id: str) -> str:
     ret = ''
+    if '[' in url_or_id:
+        url_or_id = url_or_id.replace('[', '').replace(']', '')
+
     if type(url_or_id) != str:
         raise TypeError(f"{url_or_id}: URL or VideoID must be str, but {type(url_or_id)} is passed.")
     if len(url_or_id) == YT_VIDEO_ID_LENGTH: