ArchiveBox/archivebox/extractors/__init__.py

266 lines
9.9 KiB
Python

__package__ = 'archivebox.extractors'
from typing import Callable, Optional, Dict, List, Iterable, Union, Protocol, cast
import os
import sys
from pathlib import Path
from importlib import import_module
from datetime import datetime, timezone
from django.db.models import QuerySet
from ..index.schema import ArchiveResult, Link
from ..index.sql import write_link_to_sql_index
from ..index import (
load_link_details,
write_link_details,
)
from archivebox.misc.util import enforce_types
from archivebox.misc.logging_util import (
log_archiving_started,
log_archiving_paused,
log_archiving_finished,
log_link_archiving_started,
log_link_archiving_finished,
log_archive_method_started,
log_archive_method_finished,
)
ShouldSaveFunction = Callable[[Link, Optional[Path], Optional[bool]], bool]
SaveFunction = Callable[[Link, Optional[Path], int], ArchiveResult]
ArchiveMethodEntry = tuple[str, ShouldSaveFunction, SaveFunction]
def get_default_archive_methods() -> List[ArchiveMethodEntry]:
# TODO: move to abx.pm.hook.get_EXTRACTORS()
return [
# ('favicon', should_save_favicon, save_favicon),
# ('headers', should_save_headers, save_headers),
# ('singlefile', should_save_singlefile, save_singlefile),
# ('pdf', should_save_pdf, save_pdf),
# ('screenshot', should_save_screenshot, save_screenshot),
# ('dom', should_save_dom, save_dom),
# ('wget', should_save_wget, save_wget),
# # keep title, readability, and htmltotext below wget and singlefile, as they depend on them
# ('title', should_save_title, save_title),
# ('readability', should_save_readability, save_readability),
# ('mercury', should_save_mercury, save_mercury),
# ('htmltotext', should_save_htmltotext, save_htmltotext),
# ('git', should_save_git, save_git),
# ('media', should_save_media, save_media),
# ('archive_org', should_save_archive_dot_org, save_archive_dot_org),
]
ARCHIVE_METHODS_INDEXING_PRECEDENCE = [
('readability', 1),
('mercury', 2),
('htmltotext', 3),
('singlefile', 4),
('dom', 5),
('wget', 6)
]
@enforce_types
def get_archive_methods_for_link(link: Link) -> Iterable[ArchiveMethodEntry]:
from archivebox.config.common import ARCHIVING_CONFIG
DEFAULT_METHODS = get_default_archive_methods()
allowed_methods = {
method_name
for url_pattern, methods in ARCHIVING_CONFIG.SAVE_ALLOWLIST_PTNS.items()
for method_name in methods
if url_pattern.search(link.url)
} or { method[0] for method in DEFAULT_METHODS }
denied_methods = {
method_name
for url_pattern, methods in ARCHIVING_CONFIG.SAVE_DENYLIST_PTNS.items()
for method_name in methods
if url_pattern.search(link.url)
}
allowed_methods -= denied_methods
return [method for method in DEFAULT_METHODS if method[0] in allowed_methods]
@enforce_types
def ignore_methods(to_ignore: List[str]) -> Iterable[str]:
ARCHIVE_METHODS = get_default_archive_methods()
return [method[0] for method in ARCHIVE_METHODS if method[0] not in to_ignore]
@enforce_types
def archive_link(link: Link, overwrite: bool=False, methods: Optional[Iterable[str]]=None, out_dir: Optional[Path]=None, created_by_id: int | None=None) -> Link:
"""download the DOM, PDF, and a screenshot into a folder named after the link's timestamp"""
from django.conf import settings
from ..search import write_search_index
# TODO: Remove when the input is changed to be a snapshot. Suboptimal approach.
from core.models import Snapshot, ArchiveResult
try:
snapshot = Snapshot.objects.get(url=link.url) # TODO: This will be unnecessary once everything is a snapshot
except Snapshot.DoesNotExist:
snapshot = write_link_to_sql_index(link, created_by_id=created_by_id)
active_methods = get_archive_methods_for_link(link)
if methods:
active_methods = [
method for method in active_methods
if method[0] in methods
]
out_dir = out_dir or Path(link.link_dir)
try:
is_new = not Path(out_dir).exists()
if is_new:
os.makedirs(out_dir)
link = load_link_details(link, out_dir=out_dir)
write_link_details(link, out_dir=out_dir, skip_sql_index=False)
log_link_archiving_started(link, str(out_dir), is_new)
link = link.overwrite(downloaded_at=datetime.now(timezone.utc))
stats = {'skipped': 0, 'succeeded': 0, 'failed': 0}
start_ts = datetime.now(timezone.utc)
for method_name, should_run, method_function in active_methods:
try:
if method_name not in link.history:
link.history[method_name] = []
if should_run(link, out_dir, overwrite):
log_archive_method_started(method_name)
result = method_function(link=link, out_dir=out_dir)
link.history[method_name].append(result)
stats[result.status] += 1
log_archive_method_finished(result)
write_search_index(link=link, texts=result.index_texts)
ArchiveResult.objects.create(snapshot=snapshot, extractor=method_name, cmd=result.cmd, cmd_version=result.cmd_version,
output=result.output, pwd=result.pwd, start_ts=result.start_ts, end_ts=result.end_ts, status=result.status, created_by_id=snapshot.created_by_id)
# bump the downloaded_at time on the main Snapshot here, this is critical
# to be able to cache summaries of the ArchiveResults for a given
# snapshot without having to load all the results from the DB each time.
# (we use {Snapshot.pk}-{Snapshot.downloaded_at} as the cache key and assume
# ArchiveResults are unchanged as long as the downloaded_at timestamp is unchanged)
snapshot.save()
else:
# print('{black} X {}{reset}'.format(method_name, **ANSI))
stats['skipped'] += 1
except Exception as e:
# https://github.com/ArchiveBox/ArchiveBox/issues/984#issuecomment-1150541627
with open(settings.ERROR_LOG, "a", encoding='utf-8') as f:
command = ' '.join(sys.argv)
ts = datetime.now(timezone.utc).strftime('%Y-%m-%d__%H:%M:%S')
f.write(("\n" + 'Exception in archive_methods.save_{}(Link(url={})) command={}; ts={}'.format(
method_name,
link.url,
command,
ts
) + "\n" + str(e) + "\n"))
#f.write(f"\n> {command}; ts={ts} version={config['VERSION']} docker={config['IN_DOCKER']} is_tty={config['IS_TTY']}\n")
# print(f' ERROR: {method_name} {e.__class__.__name__}: {e} {getattr(e, "hints", "")}', ts, link.url, command)
raise e from Exception('Exception in archive_methods.save_{}(Link(url={}))'.format(
method_name,
link.url,
))
# print(' ', stats)
try:
latest_title = link.history['title'][-1].output.strip()
if latest_title and len(latest_title) >= len(link.title or ''):
link = link.overwrite(title=latest_title)
except Exception:
pass
write_link_details(link, out_dir=out_dir, skip_sql_index=False)
log_link_archiving_finished(link, out_dir, is_new, stats, start_ts)
except KeyboardInterrupt:
try:
write_link_details(link, out_dir=link.link_dir)
except:
pass
raise
except Exception as err:
print(' ! Failed to archive link: {}: {}'.format(err.__class__.__name__, err))
raise
return link
@enforce_types
def archive_links(all_links: Union[Iterable[Link], QuerySet], overwrite: bool=False, methods: Optional[Iterable[str]]=None, out_dir: Optional[Path]=None, created_by_id: int | None=None) -> List[Link]:
if type(all_links) is QuerySet:
num_links: int = all_links.count()
get_link = lambda x: x.as_link_with_details()
all_links = all_links.iterator(chunk_size=500)
else:
num_links: int = len(all_links)
get_link = lambda x: x
if num_links == 0:
return []
log_archiving_started(num_links)
idx: int = 0
try:
for link in all_links:
idx += 1
to_archive = get_link(link)
archive_link(to_archive, overwrite=overwrite, methods=methods, out_dir=Path(link.link_dir), created_by_id=created_by_id)
except KeyboardInterrupt:
log_archiving_paused(num_links, idx, link.timestamp)
raise SystemExit(0)
except BaseException:
print()
raise
log_archiving_finished(num_links)
return all_links
EXTRACTORS_DIR = Path(__file__).parent
class ExtractorModuleProtocol(Protocol):
"""Type interface for an Extractor Module (WIP)"""
get_output_path: Callable
# TODO:
# get_embed_path: Callable | None
# should_extract(Snapshot)
# extract(Snapshot)
def get_extractors(dir: Path=EXTRACTORS_DIR) -> Dict[str, ExtractorModuleProtocol]:
"""iterate through archivebox/extractors/*.py and load extractor modules"""
EXTRACTORS = {}
for filename in EXTRACTORS_DIR.glob('*.py'):
if filename.name.startswith('__'):
continue
extractor_name = filename.name.replace('.py', '')
extractor_module = cast(ExtractorModuleProtocol, import_module(f'.{extractor_name}', package=__package__))
assert getattr(extractor_module, 'get_output_path')
EXTRACTORS[extractor_name] = extractor_module
return EXTRACTORS
EXTRACTORS = get_extractors(EXTRACTORS_DIR)