mirrors/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2025-05-19 01:15:11 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

Merge pull request #1021 from renaisun/dev

Browse source
This commit is contained in:
Nick Sweeting 2023-01-09 18:17:39 -08:00 • committed by GitHub
commit 0cbeeb4346
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
2 changed files with 26 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

2
archivebox/config.py
View file

@ -177,6 +177,7 @@ CONFIG_SCHEMA: Dict[str, ConfigDefaultDict] = {
'--compressed'
]},
'GIT_ARGS': {'type': list, 'default': ['--recursive']},
'SINGLEFILE_ARGS': {'type': list, 'default' : None}
},
'SEARCH_BACKEND_CONFIG' : {
@ -402,6 +403,7 @@ DYNAMIC_CONFIG_SCHEMA: ConfigDefaultDict = {
'USE_SINGLEFILE': {'default': lambda c: c['USE_SINGLEFILE'] and c['SAVE_SINGLEFILE']},
'SINGLEFILE_VERSION': {'default': lambda c: bin_version(c['SINGLEFILE_BINARY']) if c['USE_SINGLEFILE'] else None},
'SINGLEFILE_ARGS': {'default': lambda c: c['SINGLEFILE_ARGS'] or []},
'USE_READABILITY': {'default': lambda c: c['USE_READABILITY'] and c['SAVE_READABILITY']},
'READABILITY_VERSION': {'default': lambda c: bin_version(c['READABILITY_BINARY']) if c['USE_READABILITY'] else None},

26
archivebox/extractors/singlefile.py
View file

@ -17,6 +17,7 @@ from ..config import (
SAVE_SINGLEFILE,
DEPENDENCIES,
SINGLEFILE_VERSION,
SINGLEFILE_ARGS,
CHROME_BINARY,
)
from ..logging_util import TimedProgress
@ -45,10 +46,31 @@ def save_singlefile(link: Link, out_dir: Optional[Path]=None, timeout: int=TIMEO
# SingleFile CLI Docs: https://github.com/gildas-lormeau/SingleFile/tree/master/cli
browser_args = '--browser-args={}'.format(json.dumps(browser_args[1:]))
cmd = [
DEPENDENCIES['SINGLEFILE_BINARY']['path'],
options = [
*SINGLEFILE_ARGS,
'--browser-executable-path={}'.format(CHROME_BINARY),
browser_args,
]
# Deduplicate options (single-file doesn't like when you use the same option two times)
#
# NOTE: Options names that come first clobber conflicting names that come later
# My logic is SINGLEFILE_ARGS is the option that affects the singlefile command with most
# specificity, therefore the user sets it with a lot intent, therefore it should take precedence
# kind of like the ergonomic principle of lexical scope in programming languages.
seen_option_names = []
def test_seen(argument):
option_name = argument.split("=")[0]
if option_name in seen_option_names:
return False
else:
seen_option_names.append(option_name)
return True
deduped_options = list(filter(test_seen, options))
cmd = [
DEPENDENCIES['SINGLEFILE_BINARY']['path'],
*deduped_options,
link.url,
output,
]