From 911b84a79b43dabdb189056bb96e873fd0f4c6e7 Mon Sep 17 00:00:00 2001 From: Nick Sweeting Date: Fri, 11 Jan 2019 22:13:51 -0500 Subject: [PATCH] make wget cli options easier to read --- archivebox/archive_methods.py | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) diff --git a/archivebox/archive_methods.py b/archivebox/archive_methods.py index 7beaeaca..83d45520 100644 --- a/archivebox/archive_methods.py +++ b/archivebox/archive_methods.py @@ -210,8 +210,18 @@ def fetch_wget(link_dir, link, requisites=FETCH_WGET_REQUISITES, timeout=TIMEOUT CMD = [ # WGET CLI Docs: https://www.gnu.org/software/wget/manual/wget.html - *'wget -N -E -np -x -H -k -K -S --restrict-file-names=unix'.split(' '), - *(('-p',) if FETCH_WGET_REQUISITES else ()), + 'wget', + # '--server-response', + '--no-verbose', + '--timestamping', + '--adjust-extension', + '--convert-links', + '--force-directories', + '--backup-converted', + '--span-hosts', + '--no-parent', + '--restrict-file-names=unix', + *(('--page-requisites',) if FETCH_WGET_REQUISITES else ()), *(('--user-agent="{}"'.format(WGET_USER_AGENT),) if WGET_USER_AGENT else ()), *((() if CHECK_SSL_VALIDITY else ('--no-check-certificate',))), link['url'],