Merge branch 'hotfix/memory'

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__      = '0.2.1'
+__version__      = '0.2.4'
 __license__      = 'MIT'
 __author__       = 'taizan-hokuto'
 __author_email__ = '55448286+taizan-hokuto@users.noreply.github.com'

pytchat/cli/__init__.py

@@ -2,6 +2,7 @@ import argparse
 
 import os
 import signal
+import time
 from json.decoder import JSONDecodeError
 from pathlib import Path
 from .arguments import Arguments
@@ -55,7 +56,21 @@ def main():
                 path = Path(Arguments().output + video_id + '.html')
             else:
                 raise FileNotFoundError
-            info = VideoInfo(video_id)
+            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 len(Arguments().video_ids) > 1:
                 print(f"\n{'-' * 10} video:{counter + 1} of {len(Arguments().video_ids)} {'-' * 10}")
             print(f"\n"
@@ -66,7 +81,7 @@ def main():
             print(f" output path: {path.resolve()}")
             duration = info.get_duration()
             pbar = ProgressBar(total=(duration * 1000), status="Extracting")
-            ex = Extractor(video_id,               
+            ex = Extractor(video_id,
                     callback=pbar._disp,
                     div=10)
             signal.signal(signal.SIGINT, (lambda a, b: cancel(ex, pbar)))
@@ -86,8 +101,6 @@ def main():
             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:
@@ -96,14 +109,9 @@ def main():
             print("The specified directory does not exist.:{}".format(Arguments().output))
         except JSONDecodeError as e:
             print(e.msg)
-            print("Cannot parse video information.:{}".format(video_id))
+            print("JSONDecodeError.:{}".format(video_id))
             if Arguments().save_error_data:
                 util.save(e.doc, "ERR_JSON_DECODE", ".dat")
-        except PatternUnmatchError as e:
-            print(e.msg)
-            print("Cannot parse video information.:{}".format(video_id))
-            if Arguments().save_error_data:
-                util.save(e.doc, "ERR_PATTERN_UNMATCH", ".dat")
 
     return
 

pytchat/exceptions.py

@@ -38,7 +38,9 @@ class InvalidVideoIdException(Exception):
     '''
     Thrown when the video_id is not exist (VideoInfo).
     '''
-    pass
+    def __init__(self, doc):
+        self.msg = "InvalidVideoIdException"
+        self.doc = doc
 
 
 class UnknownConnectionError(Exception):
@@ -47,7 +49,7 @@ class UnknownConnectionError(Exception):
 
 class RetryExceedMaxCount(Exception):
     '''
-    thrown when the number of retries exceeds the maximum value.
+    Thrown when the number of retries exceeds the maximum value.
     '''
     pass
 
@@ -66,13 +68,13 @@ class FailedExtractContinuation(ChatDataFinished):
 
 class VideoInfoParseError(Exception):
     '''
-    thrown when failed to parse video info
+    Base exception when parsing video info.
     '''
 
 
 class PatternUnmatchError(VideoInfoParseError):
     '''
-    thrown when failed to parse video info with unmatched pattern
+    Thrown when failed to parse video info with unmatched pattern.
     '''
     def __init__(self, doc):
         self.msg = "PatternUnmatchError"

pytchat/tool/extract/asyncdl.py

@@ -16,6 +16,9 @@ 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 +53,7 @@ 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
 
@@ -64,6 +68,10 @@ def ready_blocks(video_id, duration, div, callback):
         url = f"{REPLAY_URL}{quote(continuation)}&pbj=1"
         for _ in range(MAX_RETRY_COUNT):
             try:
+                if continuation in param_set:
+                    next_continuation, actions = None, []
+                    break
+                param_set.add(continuation)
                 resp = await session.get(url, headers=headers)
                 next_continuation, actions = parser.parse(resp.json())
                 break
@@ -112,6 +120,10 @@ def fetch_patch(callback, blocks, video_id):
         url = f"{REPLAY_URL}{quote(continuation)}&pbj=1"
         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

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.