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

refactor: favicon uses snapshot instead of link

Browse source
This commit is contained in:
Cristian 2020-12-29 14:07:15 -05:00
parent d1326c3660
commit eac29e3a7a
1 changed files with 7 additions and 5 deletions
Download patch file Download diff file Expand all files Collapse all files

12
archivebox/extractors/favicon.py
View file

@ -4,6 +4,8 @@ from pathlib import Path
from typing import Optional
from django.db.models import Model
from ..index.schema import Link, ArchiveResult, ArchiveOutput
from ..system import chmod_file, run
from ..util import enforce_types, domain
@ -20,18 +22,18 @@ from ..logging_util import TimedProgress
@enforce_types
def should_save_favicon(link: Link, out_dir: Optional[str]=None) -> bool:
out_dir = out_dir or link.link_dir
def should_save_favicon(snapshot: Model, out_dir: Optional[str]=None) -> bool:
out_dir = out_dir or snapshot.snapshot_dir
if (Path(out_dir) / 'favicon.ico').exists():
return False
return SAVE_FAVICON
@enforce_types
def save_favicon(link: Link, out_dir: Optional[Path]=None, timeout: int=TIMEOUT) -> ArchiveResult:
def save_favicon(snapshot: Model, out_dir: Optional[Path]=None, timeout: int=TIMEOUT) -> ArchiveResult:
"""download site favicon from google's favicon api"""
out_dir = out_dir or link.link_dir
out_dir = out_dir or snapshot.snapshot_dir
output: ArchiveOutput = 'favicon.ico'
cmd = [
CURL_BINARY,
@ -40,7 +42,7 @@ def save_favicon(link: Link, out_dir: Optional[Path]=None, timeout: int=TIMEOUT)
'--output', str(output),
*(['--user-agent', '{}'.format(CURL_USER_AGENT)] if CURL_USER_AGENT else []),
*([] if CHECK_SSL_VALIDITY else ['--insecure']),
'https://www.google.com/s2/favicons?domain={}'.format(domain(link.url)),
'https://www.google.com/s2/favicons?domain={}'.format(domain(snapshot.url)),
]
status = 'pending'
timer = TimedProgress(timeout, prefix=' ')