mirrors/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2025-05-17 08:34:26 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

refactor: Initial and dirty refactor to replace link with snapshot. Barely functional add command

Browse source
This commit is contained in:
Cristian 2020-12-23 14:51:42 -05:00
parent 8e2270e21b
commit 8c4ae73d65
13 changed files with 246 additions and 233 deletions
Download patch file Download diff file Expand all files Collapse all files

7
archivebox/parsers/generic_html.py
View file

@ -31,6 +31,7 @@ class HrefParser(HTMLParser):
@enforce_types
def parse_generic_html_export(html_file: IO[str], root_url: Optional[str]=None, **_kwargs) -> Iterable[Link]:
"""Parse Generic HTML for href tags and use only the url (support for title coming later)"""
from core.models import Snapshot
html_file.seek(0)
for line in html_file:
@ -44,10 +45,10 @@ def parse_generic_html_export(html_file: IO[str], root_url: Optional[str]=None,
url = urljoin(root_url, url)
for archivable_url in re.findall(URL_REGEX, url):
yield Link(
yield Snapshot(
url=htmldecode(archivable_url),
timestamp=str(datetime.now().timestamp()),
title=None,
tags=None,
sources=[html_file.name],
#tags=None,
#sources=[html_file.name],
)