simon987/od-database
1
0
Fork 0
mirror of https://github.com/simon987/od-database.git synced 2025-04-20 10:56:47 +00:00
Code Issues Projects Releases Wiki Activity

Disabled thread pool for headers requests in listing

Browse Source
This commit is contained in:
Simon 2018-06-18 10:33:33 -04:00
parent b97b8f6784
commit 99d64b658b
2 changed files with 2 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
crawl_server/crawler.py
View File

@ -120,7 +120,6 @@ class RemoteDirectoryCrawler:
files_q.put(None)
file_writer_thread.join()
return CrawlResult(files_written[0], "success")
def _process_listings(self, url: str, in_q: Queue, files_q: Queue):

5
crawl_server/remote_http.py
View File

@ -1,5 +1,4 @@
from urllib.parse import urljoin, unquote, quote
from urllib.parse import unquote
import os
from html.parser import HTMLParser
from itertools import repeat
@ -104,7 +103,7 @@ class HttpDirectory(RemoteDirectory):
def request_files(self, urls_to_request: list) -> list:
if len(urls_to_request) > 30:
if len(urls_to_request) > 3000000:
# Many urls, use multi-threaded solution
pool = ThreadPool(processes=10)
files = pool.starmap(HttpDirectory._request_file, zip(repeat(self), urls_to_request))