mirrors/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2025-05-23 03:06:55 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

move almost all config into new archivebox.CONSTANTS
Some checks are pending
CodeQL / Analyze (python) (push) Waiting to run
Details
Build Debian package / build (push) Waiting to run
Details
Build Docker image / buildx (push) Waiting to run
Details
Build Homebrew package / build (push) Waiting to run
Details
Build GitHub Pages website / build (push) Waiting to run
Details
Build GitHub Pages website / deploy (push) Blocked by required conditions
Details
Run linters / lint (push) Waiting to run
Details
Build Pip package / build (push) Waiting to run
Details
Run tests / python_tests (ubuntu-22.04, 3.11) (push) Waiting to run
Details
Run tests / docker_tests (push) Waiting to run
Details

Browse source
This commit is contained in:
Nick Sweeting 2024-09-25 05:10:09 -07:00
parent f5e8d99fdf
commit bb65b2dbec
No known key found for this signature in database
32 changed files with 982 additions and 840 deletions
Download patch file Download diff file Expand all files Collapse all files

5
archivebox/extractors/htmltotext.py
View file

@ -1,5 +1,7 @@
__package__ = 'archivebox.extractors'
import archivebox
from html.parser import HTMLParser
import io
from pathlib import Path
@ -8,7 +10,6 @@ from typing import Optional
from ..config import (
SAVE_HTMLTOTEXT,
TIMEOUT,
VERSION,
)
from ..index.schema import Link, ArchiveResult, ArchiveError
from ..logging_util import TimedProgress
@ -153,7 +154,7 @@ def save_htmltotext(link: Link, out_dir: Optional[Path]=None, timeout: int=TIMEO
return ArchiveResult(
cmd=cmd,
pwd=str(out_dir),
cmd_version=VERSION,
cmd_version=archivebox.__version__,
output=output,
status=status,
index_texts=[extracted_text] if extracted_text else [],