Mirrors/ArchiveBox
2
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox synced 2024-11-10 06:34:16 +00:00
Code Issues Projects Releases Packages Wiki Activity

feat: Add CURL_ARGS to control curl arguments

Browse source
This commit is contained in:
Cristian 2020-10-15 08:42:46 -05:00 committed by Cristian Vargas
parent 24e7a74855
commit 972d57bd08
2 changed files with 9 additions and 5 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

8
archivebox/config/__init__.py
View file

@ -130,7 +130,12 @@ CONFIG_DEFAULTS: Dict[str, ConfigDefaultDict] = {
'--span-hosts',
'--no-parent',
'-e', 'robots=off',
]}
]},
'CURL_ARGS': {'type': list, 'default': ['--silent',
'--location',
'--head',
'--compressed'
]}
},
'DEPENDENCY_CONFIG': {
@ -277,6 +282,7 @@ DERIVED_CONFIG_DEFAULTS: ConfigDefaultDict = {
'USE_CURL': {'default': lambda c: c['USE_CURL'] and (c['SAVE_FAVICON'] or c['SAVE_TITLE'] or c['SAVE_ARCHIVE_DOT_ORG'])},
'CURL_VERSION': {'default': lambda c: bin_version(c['CURL_BINARY']) if c['USE_CURL'] else None},
'CURL_USER_AGENT': {'default': lambda c: c['CURL_USER_AGENT'].format(**c)},
'CURL_ARGS': {'default': lambda c: c['CURL_ARGS'] or []},
'SAVE_FAVICON': {'default': lambda c: c['USE_CURL'] and c['SAVE_FAVICON']},
'SAVE_ARCHIVE_DOT_ORG': {'default': lambda c: c['USE_CURL'] and c['SAVE_ARCHIVE_DOT_ORG']},

6
archivebox/extractors/archive_org.py
View file

@ -13,6 +13,7 @@ from ..util import (
)
from ..config import (
TIMEOUT,
CURL_ARGS,
CHECK_SSL_VALIDITY,
SAVE_ARCHIVE_DOT_ORG,
CURL_BINARY,
@ -45,10 +46,7 @@ def save_archive_dot_org(link: Link, out_dir: Optional[Path]=None, timeout: int=
submit_url = 'https://web.archive.org/save/{}'.format(link.url)
cmd = [
CURL_BINARY,
'--silent',
'--location',
'--head',
'--compressed',
*CURL_ARGS,
'--max-time', str(timeout),
*(['--user-agent', '{}'.format(CURL_USER_AGENT)] if CURL_USER_AGENT else []),
*([] if CHECK_SSL_VALIDITY else ['--insecure']),