mirror of
https://github.com/ArchiveBox/ArchiveBox.git
synced 2025-05-22 10:55:18 -04:00
refactor: singlefile uses snapshot instead of link
This commit is contained in:
parent
f6152ded44
commit
ca7f48042b
3 changed files with 11 additions and 9 deletions
|
@ -5,7 +5,9 @@ from pathlib import Path
|
||||||
from typing import Optional
|
from typing import Optional
|
||||||
import json
|
import json
|
||||||
|
|
||||||
from ..index.schema import Link, ArchiveResult, ArchiveError
|
from django.db.models import Model
|
||||||
|
|
||||||
|
from ..index.schema import ArchiveResult, ArchiveError
|
||||||
from ..system import run, chmod_file
|
from ..system import run, chmod_file
|
||||||
from ..util import (
|
from ..util import (
|
||||||
enforce_types,
|
enforce_types,
|
||||||
|
@ -23,9 +25,9 @@ from ..logging_util import TimedProgress
|
||||||
|
|
||||||
|
|
||||||
@enforce_types
|
@enforce_types
|
||||||
def should_save_singlefile(link: Link, out_dir: Optional[Path]=None) -> bool:
|
def should_save_singlefile(snapshot: Model, out_dir: Optional[Path]=None) -> bool:
|
||||||
out_dir = out_dir or Path(link.link_dir)
|
out_dir = out_dir or Path(snapshot.snapshot_dir)
|
||||||
if is_static_file(link.url):
|
if is_static_file(snapshot.url):
|
||||||
return False
|
return False
|
||||||
|
|
||||||
output = out_dir / 'singlefile.html'
|
output = out_dir / 'singlefile.html'
|
||||||
|
@ -33,10 +35,10 @@ def should_save_singlefile(link: Link, out_dir: Optional[Path]=None) -> bool:
|
||||||
|
|
||||||
|
|
||||||
@enforce_types
|
@enforce_types
|
||||||
def save_singlefile(link: Link, out_dir: Optional[Path]=None, timeout: int=TIMEOUT) -> ArchiveResult:
|
def save_singlefile(snapshot: Model, out_dir: Optional[Path]=None, timeout: int=TIMEOUT) -> ArchiveResult:
|
||||||
"""download full site using single-file"""
|
"""download full site using single-file"""
|
||||||
|
|
||||||
out_dir = out_dir or Path(link.link_dir)
|
out_dir = out_dir or Path(snapshot.snapshot_dir)
|
||||||
output = str(out_dir.absolute() / "singlefile.html")
|
output = str(out_dir.absolute() / "singlefile.html")
|
||||||
|
|
||||||
browser_args = chrome_args(TIMEOUT=0)
|
browser_args = chrome_args(TIMEOUT=0)
|
||||||
|
@ -47,7 +49,7 @@ def save_singlefile(link: Link, out_dir: Optional[Path]=None, timeout: int=TIMEO
|
||||||
DEPENDENCIES['SINGLEFILE_BINARY']['path'],
|
DEPENDENCIES['SINGLEFILE_BINARY']['path'],
|
||||||
'--browser-executable-path={}'.format(CHROME_BINARY),
|
'--browser-executable-path={}'.format(CHROME_BINARY),
|
||||||
browser_args,
|
browser_args,
|
||||||
link.url,
|
snapshot.url,
|
||||||
output
|
output
|
||||||
]
|
]
|
||||||
|
|
||||||
|
|
|
@ -8,7 +8,7 @@ from datetime import datetime
|
||||||
|
|
||||||
from django.db.models import Model
|
from django.db.models import Model
|
||||||
|
|
||||||
from ..index.schema import Link, ArchiveResult, ArchiveOutput, ArchiveError
|
from ..index.schema import ArchiveResult, ArchiveOutput, ArchiveError
|
||||||
from ..system import run, chmod_file
|
from ..system import run, chmod_file
|
||||||
from ..util import (
|
from ..util import (
|
||||||
enforce_types,
|
enforce_types,
|
||||||
|
|
2
package-lock.json
generated
2
package-lock.json
generated
|
@ -1,6 +1,6 @@
|
||||||
{
|
{
|
||||||
"name": "archivebox",
|
"name": "archivebox",
|
||||||
"version": "0.4.21",
|
"version": "0.5.1",
|
||||||
"lockfileVersion": 1,
|
"lockfileVersion": 1,
|
||||||
"requires": true,
|
"requires": true,
|
||||||
"dependencies": {
|
"dependencies": {
|
||||||
|
|
Loading…
Add table
Add a link
Reference in a new issue