new vastly simplified plugin spec without pydantic
Some checks are pending
Build Debian package / build (push) Waiting to run
Build Docker image / buildx (push) Waiting to run
Build Homebrew package / build (push) Waiting to run
Run linters / lint (push) Waiting to run
Build Pip package / build (push) Waiting to run
Run tests / python_tests (ubuntu-22.04, 3.11) (push) Waiting to run
Run tests / docker_tests (push) Waiting to run

This commit is contained in:
Nick Sweeting 2024-10-14 21:50:47 -07:00
parent abf75f49f4
commit 01ba6d49d3
No known key found for this signature in database
115 changed files with 2466 additions and 2301 deletions

View file

@ -6,12 +6,16 @@ from tempfile import NamedTemporaryFile
from typing import Optional
import json
from ..index.schema import Link, ArchiveResult, ArchiveError
from archivebox.misc.system import run, atomic_write
from archivebox.misc.util import enforce_types, is_static_file
from ..index.schema import Link, ArchiveResult, ArchiveError
from ..logging_util import TimedProgress
from .title import get_html
from plugins_extractor.readability.config import READABILITY_CONFIG
from plugins_extractor.readability.binaries import READABILITY_BINARY
def get_output_path():
return 'readability/'
@ -21,7 +25,6 @@ def get_embed_path(archiveresult=None):
@enforce_types
def should_save_readability(link: Link, out_dir: Optional[str]=None, overwrite: Optional[bool]=False) -> bool:
from plugins_extractor.readability.apps import READABILITY_CONFIG
if is_static_file(link.url):
return False
@ -37,8 +40,6 @@ def should_save_readability(link: Link, out_dir: Optional[str]=None, overwrite:
def save_readability(link: Link, out_dir: Optional[str]=None, timeout: int=0) -> ArchiveResult:
"""download reader friendly version using @mozilla/readability"""
from plugins_extractor.readability.apps import READABILITY_CONFIG, READABILITY_BINARY
READABILITY_BIN = READABILITY_BINARY.load()
assert READABILITY_BIN.abspath and READABILITY_BIN.version