mirrors/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2025-05-15 15:44:26 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

better Link string and is_archived methods

Browse source
This commit is contained in:
Nick Sweeting 2019-04-24 11:36:47 -04:00
parent 11fd436305
commit 8f63f24002
1 changed files with 16 additions and 6 deletions
Download patch file Download diff file Expand all files Collapse all files

22
archivebox/legacy/schema.py
View file

@ -92,6 +92,9 @@ class Link:
updated: Optional[datetime] = None updated: Optional[datetime] = None
schema: str = 'Link' schema: str = 'Link'
def __str__(self) -> str:
return f'[{self.timestamp}] {self.base_url} "{self.title}"'
def __post_init__(self): def __post_init__(self):
self.typecheck() self.typecheck()
@ -215,8 +218,8 @@ class Link:
@property @property
def link_dir(self) -> str: def link_dir(self) -> str:
from .config import ARCHIVE_DIR from .config import CONFIG
return os.path.join(ARCHIVE_DIR, self.timestamp) return os.path.join(CONFIG['ARCHIVE_DIR'], self.timestamp)
@property @property
def archive_path(self) -> str: def archive_path(self) -> str:
@ -309,11 +312,18 @@ class Link:
from .config import ARCHIVE_DIR from .config import ARCHIVE_DIR
from .util import domain from .util import domain
return os.path.exists(os.path.join( output_paths = (
ARCHIVE_DIR,
self.timestamp,
domain(self.url), domain(self.url),
)) 'output.pdf',
'screenshot.png',
'output.html',
'media',
)
return any(
os.path.exists(os.path.join(ARCHIVE_DIR, self.timestamp, path))
for path in output_paths
)
def latest_outputs(self, status: str=None) -> Dict[str, ArchiveOutput]: def latest_outputs(self, status: str=None) -> Dict[str, ArchiveOutput]:
"""get the latest output that each archive method produced for link""" """get the latest output that each archive method produced for link"""