remove debug print

nhentai/command.py

@@ -77,7 +77,7 @@ def main():
         doujinshi_ids = list(set(map(int, doujinshi_ids)) - set(data))
 
     if not options.is_show:
-        downloader = Downloader(path=options.output_dir, threads=options.threads,
+        downloader = Downloader(path=options.output_dir, size=options.threads,
                                 timeout=options.timeout, delay=options.delay)
 
         for doujinshi_id in doujinshi_ids:

nhentai/downloader.py

@@ -1,17 +1,24 @@
 # coding: utf-
 
+import multiprocessing
+import signal
+
+import sys
 import os
-import asyncio
-import httpx
+import requests
+import time
 import urllib3.exceptions
 
 from urllib.parse import urlparse
 from nhentai import constant
 from nhentai.logger import logger
-from nhentai.utils import Singleton, async_request
+from nhentai.parser import request
+from nhentai.utils import Singleton
 
 
 urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning)
+semaphore = multiprocessing.Semaphore(1)
+
 
 class NHentaiImageNotExistException(Exception):
     pass
@@ -32,66 +39,64 @@ def download_callback(result):
         logger.log(16, f'{data} downloaded successfully')
 
 
-async def fiber(tasks):
-    for completed_task in asyncio.as_completed(tasks):
-        try:
-            result = await completed_task
-            logger.info(f'{result[1]} download completed')
-        except Exception as e:
-            logger.error(f'An error occurred: {e}')
-
-
 class Downloader(Singleton):
-    def __init__(self, path='', threads=5, timeout=30, delay=0):
-        self.threads = threads
+
+    def __init__(self, path='', size=5, timeout=30, delay=0):
+        self.size = size
         self.path = str(path)
         self.timeout = timeout
         self.delay = delay
 
-    async def _semaphore_download(self, semaphore, *args, **kwargs):
-        async with semaphore:
-            return await self.download(*args, **kwargs)
-
-    async def download(self, url, folder='', filename='', retried=0, proxy=None):
-        logger.info(f'Starting to download {url} ...')
-
+    def download(self, url, folder='', filename='', retried=0, proxy=None):
         if self.delay:
-            await asyncio.sleep(self.delay)
-
+            time.sleep(self.delay)
+        logger.info(f'Starting to download {url} ...')
         filename = filename if filename else os.path.basename(urlparse(url).path)
+        base_filename, extension = os.path.splitext(filename)
 
-        save_file_path = os.path.join(self.folder, filename)
-
+        save_file_path = os.path.join(folder, base_filename.zfill(3) + extension)
         try:
             if os.path.exists(save_file_path):
                 logger.warning(f'Skipped download: {save_file_path} already exists')
                 return 1, url
 
-            response = await async_request('GET', url, timeout=self.timeout, proxies=proxy)
-
+            response = None
+            with open(save_file_path, "wb") as f:
+                i = 0
+                while i < 10:
+                    try:
+                        response = request('get', url, stream=True, timeout=self.timeout, proxies=proxy)
                         if response.status_code != 200:
                             path = urlparse(url).path
                             for mirror in constant.IMAGE_URL_MIRRORS:
-                    logger.info(f"Try mirror: {mirror}{path}")
+                                # print(f'{mirror}{path}')
                                 mirror_url = f'{mirror}{path}'
-                    response = await async_request('GET', mirror_url, timeout=self.timeout, proxies=proxy)
+                                response = request('get', mirror_url, stream=True,
+                                                   timeout=self.timeout, proxies=proxy)
                                 if response.status_code == 200:
                                     break
 
-            if not await self.save(filename, response):
-                logger.error(f'Can not download image {url}')
-                return 1, None
+                    except Exception as e:
+                        i += 1
+                        if not i < 10:
+                            logger.critical(str(e))
+                            return 0, None
+                        continue
 
-        except (httpx.HTTPStatusError, httpx.TimeoutException, httpx.ConnectError) as e:
+                    break
+
+                length = response.headers.get('content-length')
+                if length is None:
+                    f.write(response.content)
+                else:
+                    for chunk in response.iter_content(2048):
+                        f.write(chunk)
+
+        except (requests.HTTPError, requests.Timeout) as e:
             if retried < 3:
-                logger.info(f'Download {filename} failed, retrying({retried + 1}) times...')
-                return await self.download(
-                    url=url,
-                    folder=folder,
-                    filename=filename,
-                    retried=retried + 1,
-                    proxy=proxy,
-                )
+                logger.warning(f'Warning: {e}, retrying({retried}) ...')
+                return 0, self.download(url=url, folder=folder, filename=filename,
+                                        retried=retried+1, proxy=proxy)
             else:
                 return 0, None
 
@@ -101,8 +106,6 @@ class Downloader(Singleton):
 
         except Exception as e:
             import traceback
-
-            logger.error(f"Exception type: {type(e)}")
             traceback.print_stack()
             logger.critical(str(e))
             return 0, None
@@ -112,22 +115,6 @@ class Downloader(Singleton):
 
         return 1, url
 
-    async def save(self, save_file_path, response) -> bool:
-        if response is None:
-            logger.error('Error: Response is None')
-            return False
-        save_file_path = os.path.join(self.folder, save_file_path)
-        with open(save_file_path, 'wb') as f:
-            if response is not None:
-                length = response.headers.get('content-length')
-                if length is None:
-                    f.write(response.content)
-                else:
-                    async for chunk in response.aiter_bytes(2048):
-                        f.write(chunk)
-        return True
-
-
     def start_download(self, queue, folder='') -> bool:
         if not isinstance(folder, (str, )):
             folder = str(folder)
@@ -141,20 +128,34 @@ class Downloader(Singleton):
                 os.makedirs(folder)
             except EnvironmentError as e:
                 logger.critical(str(e))
-        self.folder = folder
 
         if os.getenv('DEBUG', None) == 'NODOWNLOAD':
             # Assuming we want to continue with rest of process.
             return True
+        queue = [(self, url, folder, constant.CONFIG['proxy']) for url in queue]
 
-        semaphore = asyncio.Semaphore(self.threads)
+        pool = multiprocessing.Pool(self.size, init_worker)
+        [pool.apply_async(download_wrapper, args=item) for item in queue]
 
-        coroutines = [
-            self._semaphore_download(semaphore, url, filename=os.path.basename(urlparse(url).path))
-            for url in queue
-        ]
-
-        # Prevent coroutines infection
-        asyncio.run(fiber(coroutines))
+        pool.close()
+        pool.join()
 
         return True
+
+
+def download_wrapper(obj, url, folder='', proxy=None):
+    if sys.platform == 'darwin' or semaphore.get_value():
+        return Downloader.download(obj, url=url, folder=folder, proxy=proxy)
+    else:
+        return -3, None
+
+
+def init_worker():
+    signal.signal(signal.SIGINT, subprocess_signal)
+
+
+def subprocess_signal(sig, frame):
+    if semaphore.acquire(timeout=1):
+        logger.warning('Ctrl-C pressed, exiting sub processes ...')
+
+    raise KeyboardInterrupt

nhentai/utils.py

@@ -6,7 +6,6 @@ import os
 import zipfile
 import shutil
 
-import httpx
 import requests
 import sqlite3
 import urllib.parse
@@ -33,28 +32,8 @@ def request(method, url, **kwargs):
     return getattr(session, method)(url, verify=False, **kwargs)
 
 
-async def async_request(method, url, proxies = None, **kwargs):
-    headers = {
-        'Referer': constant.LOGIN_URL,
-        'User-Agent': constant.CONFIG['useragent'],
-        'Cookie': constant.CONFIG['cookie'],
-    }
-
-    if proxies is None:
-        proxies = constant.CONFIG['proxy']
-
-    if proxies.get('http') == '' and proxies.get('https') == '':
-        proxies = None
-
-    async with httpx.AsyncClient(headers=headers, verify=False, proxies=proxies, **kwargs) as client:
-        response = await client.request(method, url, **kwargs)
-
-    return response
-
-
 def check_cookie():
     response = request('get', constant.BASE_URL)
-
     if response.status_code == 403 and 'Just a moment...' in response.text:
         logger.error('Blocked by Cloudflare captcha, please set your cookie and useragent')
         sys.exit(1)

requirements.txt

@@ -1,4 +1,3 @@
-httpx
 requests
 soupsieve
 setuptools

tests/test_download.py

@@ -20,7 +20,7 @@ class TestDownload(unittest.TestCase):
     def test_download(self):
         did = 440546
         info = Doujinshi(**doujinshi_parser(did), name_format='%i')
-        info.downloader = Downloader(path='/tmp', threads=5)
+        info.downloader = Downloader(path='/tmp', size=5)
         info.download()
 
         self.assertTrue(os.path.exists(f'/tmp/{did}/001.jpg'))