246 lines
8.9 KiB
Python
246 lines
8.9 KiB
Python
# This file is part of the Open Parts Database software
|
|
# Copyright (C) 2022 Valentin Lorentz
|
|
#
|
|
# This program is free software: you can redistribute it and/or modify it under the
|
|
# terms of the GNU Affero General Public License version 3, as published by the
|
|
# Free Software Foundation.
|
|
#
|
|
# This program is distributed in the hope that it will be useful, but WITHOUT ANY
|
|
# WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A
|
|
# PARTICULAR PURPOSE. See the GNU Affero General Public License for more details.
|
|
#
|
|
# You should have received a copy of the GNU Affero General Public License along with
|
|
# this program. If not, see <http://www.gnu.org/licenses/>.
|
|
|
|
"""
|
|
Management of the cache of external pages.
|
|
|
|
This package fetches web pages, archives them in the Internet Archive for future
|
|
citation, and caches them in the local database for quick access by other workers.
|
|
"""
|
|
|
|
import datetime
|
|
import re
|
|
import socket
|
|
import time
|
|
import typing
|
|
|
|
import pkg_resources
|
|
import requests
|
|
|
|
from opdb.db import Db, models
|
|
|
|
_OPDB_VERSION = pkg_resources.require("opdb")[0].version
|
|
USER_AGENT = (
|
|
f"OPDB/{_OPDB_VERSION} (Open Parts Database cacher; +https://git.tf/opdb/opdb)"
|
|
)
|
|
|
|
_wayback_url_re = re.compile(
|
|
r"^https?://web\.archive\.org/web/(?P<timestamp>[0-9]{14})/(?P<origin_url>.+)$"
|
|
)
|
|
|
|
|
|
def _datetime_from_ia_timestamp(ia_timestamp: str) -> datetime.datetime:
|
|
"""
|
|
>>> _datetime_from_ia_timestamp("20220919233014")
|
|
datetime.datetime(2022, 9, 19, 23, 30, 14, tzinfo=datetime.timezone.utc)
|
|
"""
|
|
dt = datetime.datetime.strptime(ia_timestamp, "%Y%m%d%H%M%S")
|
|
# Assume it's UTC (neither the Wayback API nor the documentation mention
|
|
# timezones)
|
|
return dt.replace(tzinfo=datetime.timezone.utc)
|
|
|
|
|
|
class Session:
|
|
"""
|
|
Wrapper for :class:`requests.Session`, which tries to use pages cached locally in
|
|
the postgresql database, and falls back to downloading; making sure they are
|
|
archived in the Internet Archive.
|
|
"""
|
|
|
|
def __init__(
|
|
self, db: Db, min_snapshot_date: datetime.datetime, ias3_auth: str = None
|
|
):
|
|
self.min_snapshot_date = min_snapshot_date
|
|
self._db = db
|
|
self._session = requests.Session()
|
|
self._session.headers["User-Agent"] = USER_AGENT
|
|
self._ias3_auth = ias3_auth
|
|
|
|
def _fetch_newest_wayback_snapshot(
|
|
self, url: str
|
|
) -> typing.Optional[models.WebPageSnapshot]:
|
|
"""
|
|
If the URL is already archived in the Internet Archive (and newer than
|
|
configured with ``min_snapshot_date``), retrieves the latest snapshot available
|
|
via the Wayback Machine and returns it.
|
|
"""
|
|
# API documentation: https://archive.org/help/wayback_api.php
|
|
response = self._session.get(
|
|
"https://archive.org/wayback/available", params={"url": url}
|
|
)
|
|
response.raise_for_status() # TODO: retry
|
|
|
|
newest_ia_snapshot = (
|
|
response.json().get("archived_snapshots", {}).get("closest", {})
|
|
)
|
|
if not newest_ia_snapshot:
|
|
return None
|
|
|
|
ia_timestamp = newest_ia_snapshot["timestamp"]
|
|
snapshot_date = _datetime_from_ia_timestamp(ia_timestamp)
|
|
|
|
if snapshot_date < self.min_snapshot_date:
|
|
return None
|
|
|
|
wayback_url = newest_ia_snapshot["url"]
|
|
m = _wayback_url_re.match(wayback_url)
|
|
assert m, f"Unexpected Wayback URL format: {wayback_url}"
|
|
assert ia_timestamp == m.group(
|
|
"timestamp"
|
|
), "Timestamp unexpectedly missing from snapshot URL: {wayback_url}"
|
|
|
|
return self._fetch_wayback_snapshot(url, wayback_url)
|
|
|
|
def _fetch_wayback_snapshot(
|
|
self, url: str, wayback_url: str
|
|
) -> models.WebPageSnapshot:
|
|
# Add "id_" after the timestamp in the Wayback URL; it allows fetching the
|
|
# original page without the navigation header added by the Wayback Machine.
|
|
# Documented at https://archive.org/post/1044859/
|
|
m = _wayback_url_re.match(wayback_url)
|
|
assert m, f"Unexpected Wayback URL format: {wayback_url}"
|
|
ia_timestamp = m.group("timestamp")
|
|
snapshot_url = wayback_url.replace(ia_timestamp, ia_timestamp + "id_", 1)
|
|
|
|
response = self._session.get(snapshot_url)
|
|
response.raise_for_status() # TODO: retry
|
|
|
|
return models.WebPageSnapshot(
|
|
url=url,
|
|
snapshot_date=_datetime_from_ia_timestamp(ia_timestamp),
|
|
snapshot_url=snapshot_url,
|
|
retrieved_at=datetime.datetime.now(tz=datetime.timezone.utc),
|
|
retrieved_by=socket.getfqdn(),
|
|
response_headers=dict(response.headers),
|
|
content=response.content,
|
|
)
|
|
|
|
def _save_page_now(self, url: str) -> typing.Optional[models.WebPageSnapshot]:
|
|
if self._ias3_auth is None:
|
|
return self._anonymous_save_page_now(url)
|
|
else:
|
|
return self._authenticated_save_page_now(url, self._ias3_auth)
|
|
|
|
def _anonymous_save_page_now(
|
|
self, url: str
|
|
) -> typing.Optional[models.WebPageSnapshot]:
|
|
while True:
|
|
try:
|
|
response = self._session.get(
|
|
f"https://web.archive.org/save/{url}",
|
|
allow_redirects=False,
|
|
)
|
|
response.raise_for_status()
|
|
except requests.exceptions.HTTPError as e:
|
|
if e.response.status_code == 429:
|
|
print(e)
|
|
print("Sleeping...")
|
|
time.sleep(10)
|
|
continue
|
|
elif e.response.status_code == 520:
|
|
# "Job failed". We will try again in the next workflow run.
|
|
return None
|
|
else:
|
|
raise
|
|
else:
|
|
wayback_url = response.headers["Location"]
|
|
return self._fetch_wayback_snapshot(url, wayback_url)
|
|
|
|
def _authenticated_save_page_now(
|
|
self, url: str, ias3_auth: str
|
|
) -> typing.Optional[models.WebPageSnapshot]:
|
|
for _ in range(3):
|
|
response = self._session.post(
|
|
"https://web.archive.org/save/",
|
|
allow_redirects=False,
|
|
data={"url": url},
|
|
headers={
|
|
"Accept": "application/json",
|
|
"Authorization": f"LOW {ias3_auth}",
|
|
},
|
|
)
|
|
response.raise_for_status()
|
|
if response.json().get("status_ext") == "error:too-many-daily-captures":
|
|
# typically happens when a page repeatedly fails so we
|
|
# (unsuccessfully) tried to capture it too many times
|
|
return None
|
|
job_id = response.json()["job_id"]
|
|
|
|
status = "pending"
|
|
while status == "pending":
|
|
time.sleep(5)
|
|
response = self._session.get(
|
|
f"https://web.archive.org/save/status/{job_id}"
|
|
)
|
|
response.raise_for_status()
|
|
status = response.json()["status"]
|
|
|
|
if status == "success":
|
|
break
|
|
|
|
if response.json()["status"] == "error":
|
|
print(response.json()["message"])
|
|
time.sleep(10)
|
|
continue # retry
|
|
|
|
assert False, response.json()
|
|
else:
|
|
print("Too many failures; giving up.")
|
|
return None
|
|
|
|
ia_timestamp = response.json()["timestamp"]
|
|
snapshot_date = _datetime_from_ia_timestamp(ia_timestamp)
|
|
|
|
assert snapshot_date >= self.min_snapshot_date, (
|
|
snapshot_date.isoformat(),
|
|
self.min_snapshot_date.isoformat(),
|
|
)
|
|
|
|
wayback_url = f"https://web.archive.org/web/{ia_timestamp}/{url}"
|
|
return self._fetch_wayback_snapshot(url, wayback_url)
|
|
|
|
def _get_cached_snapshot(self, url: str) -> typing.Optional[models.WebPageSnapshot]:
|
|
snapshot = self._db.get_last_web_page_snapshot(url)
|
|
if snapshot is None or snapshot.snapshot_date < self.min_snapshot_date:
|
|
return None
|
|
|
|
return snapshot
|
|
|
|
def get_or_fetch_snapshot(
|
|
self, url: str
|
|
) -> typing.Optional[models.WebPageSnapshot]:
|
|
"""
|
|
Fetches the given URL from the local cache or from the Wayback Machine.
|
|
|
|
Requests archival by the Internet Archive if the Wayback Machine does not
|
|
already have the page available.
|
|
"""
|
|
# First, try the local cache
|
|
snapshot = self._get_cached_snapshot(url)
|
|
if snapshot is not None:
|
|
return snapshot
|
|
|
|
# Then, try fetching from the Wayback Machine (and cache it locally)
|
|
snapshot = self._fetch_newest_wayback_snapshot(url)
|
|
if snapshot is not None:
|
|
self._db.add_web_page_snapshots([snapshot])
|
|
return snapshot
|
|
|
|
# If the Internet Archive does not have it yet, trigger its Save Code Now,
|
|
# and query the Wayback Machine again
|
|
snapshot = self._save_page_now(url)
|
|
if snapshot is not None:
|
|
self._db.add_web_page_snapshots([snapshot])
|
|
return snapshot
|