comictagger/comictalker/comiccacher.py
Timmy Welch f72ebdb149 Simplify ComicCacher to store a single binary data field and ID(s)
If the ComicCacher is to be a generic cache for talkers it must assume
 very little. Current assumptions:
 - There are issues that can be queried individually by an "Issue ID" and they have a relation to a single series
 - There are series that can be queried individually by an "Series ID" and they have a relation to zero or more issues
 - There are Searches that can be queried by the search term and they have a relation to zero or more series

Each series and issue have a boolean `complete` attribute which is up to the talker to decide what it means.
Data is returned as a tuple ([series, complete] or [issue, complete]) or a list of tuples
An issue consists of an ID, an series ID and a binary data attribute which is up to the talker to determine what it means.
An series consists of in ID and a binary data attribute which is up to the talker to determine what it means.

The data attribute is binary to allow for compression and efficient storage of binary data (e.g. pickle) it is suggested to store it as json or similar text format encoded with utf-8. If the talker is using a website API it is suggested to store the raw response from the server.

All caches automatically expire 7 days after insertion.
2023-08-05 03:02:12 -07:00

300 lines
9.7 KiB
Python

"""A python class to manage caching of metadata from comic sources"""
#
# Copyright 2012-2014 ComicTagger Authors
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from __future__ import annotations
import datetime
import logging
import os
import pathlib
import sqlite3
from typing import Any
from typing_extensions import NamedTuple
logger = logging.getLogger(__name__)
class Series(NamedTuple):
id: str
data: bytes
class Issue(NamedTuple):
id: str
series_id: str
data: bytes
class ComicCacher:
def __init__(self, cache_folder: pathlib.Path, version: str) -> None:
self.cache_folder = cache_folder
self.db_file = cache_folder / "comic_cache.db"
self.version_file = cache_folder / "cache_version.txt"
self.version = version
# verify that cache is from same version as this one
data = ""
try:
with open(self.version_file, "rb") as f:
data = f.read().decode("utf-8")
f.close()
except Exception:
pass
if data != version:
self.clear_cache()
if not os.path.exists(self.db_file):
self.create_cache_db()
def clear_cache(self) -> None:
try:
os.unlink(self.db_file)
except Exception:
pass
try:
os.unlink(self.version_file)
except Exception:
pass
def create_cache_db(self) -> None:
# create the version file
with open(self.version_file, "w", encoding="utf-8") as f:
f.write(self.version)
# this will wipe out any existing version
open(self.db_file, "wb").close()
con = sqlite3.connect(self.db_file)
con.row_factory = sqlite3.Row
# create tables
with con:
cur = con.cursor()
cur.execute(
"""CREATE TABLE SeriesSearchCache(
timestamp DATE DEFAULT (datetime('now','localtime')),
id TEXT NOT NULL,
source TEXT NOT NULL,
search_term TEXT,
PRIMARY KEY (id, source, search_term))"""
)
cur.execute("CREATE TABLE Source(id TEXT NOT NULL, name TEXT NOT NULL, PRIMARY KEY (id))")
cur.execute(
"""CREATE TABLE Series(
timestamp DATE DEFAULT (datetime('now','localtime')),
id TEXT NOT NULL,
source TEXT NOT NULL,
data BLOB,
complete BOOL,
PRIMARY KEY (id, source))"""
)
cur.execute(
"""CREATE TABLE Issues(
timestamp DATE DEFAULT (datetime('now','localtime')),
id TEXT NOT NULL,
source TEXT NOT NULL,
series_id TEXT,
data BLOB,
complete BOOL,
PRIMARY KEY (id, source))"""
)
def expire_stale_records(self, cur: sqlite3.Cursor, table: str) -> None:
# purge stale series info
a_week_ago = datetime.datetime.today() - datetime.timedelta(days=7)
cur.execute("DELETE FROM Series WHERE timestamp < ?", [str(a_week_ago)])
def add_search_results(self, source: str, search_term: str, series_list: list[Series], complete: bool) -> None:
with sqlite3.connect(self.db_file) as con:
con.row_factory = sqlite3.Row
con.text_factory = str
cur = con.cursor()
# remove all previous entries with this search term
cur.execute(
"DELETE FROM SeriesSearchCache WHERE search_term = ? AND source = ?",
[search_term.casefold(), source],
)
# now add in new results
for series in series_list:
cur.execute(
"INSERT INTO SeriesSearchCache (source, search_term, id) VALUES(?, ?, ?)",
(source, search_term.casefold(), series.id),
)
data = {
"id": series.id,
"source": source,
"data": series.data,
"complete": complete,
}
self.upsert(cur, "series", data)
def add_series_info(self, source: str, series: Series, complete: bool) -> None:
with sqlite3.connect(self.db_file) as con:
con.row_factory = sqlite3.Row
cur = con.cursor()
data = {
"id": series.id,
"source": source,
"data": series.data,
"complete": complete,
}
self.upsert(cur, "series", data)
def add_issues_info(self, source: str, issues: list[Issue], complete: bool) -> None:
with sqlite3.connect(self.db_file) as con:
con.row_factory = sqlite3.Row
cur = con.cursor()
for issue in issues:
data = {
"id": issue.id,
"series_id": issue.series_id,
"data": issue.data,
"source": source,
"complete": complete,
}
self.upsert(cur, "issues", data)
def get_search_results(self, source: str, search_term: str, expire_stale: bool = True) -> list[tuple[Series, bool]]:
results = []
with sqlite3.connect(self.db_file) as con:
con.row_factory = sqlite3.Row
con.text_factory = str
cur = con.cursor()
if expire_stale:
self.expire_stale_records(cur, "SeriesSearchCache")
self.expire_stale_records(cur, "Series")
cur.execute(
"""SELECT * FROM SeriesSearchCache INNER JOIN Series on
SeriesSearchCache.id=Series.id AND SeriesSearchCache.source=Series.source
WHERE search_term=? AND SeriesSearchCache.source=?""",
[search_term.casefold(), source],
)
rows = cur.fetchall()
for record in rows:
result = Series(id=record["id"], data=record["data"])
results.append((result, record["complete"]))
return results
def get_series_info(self, series_id: str, source: str, expire_stale: bool = True) -> tuple[Series, bool] | None:
result: Series | None = None
with sqlite3.connect(self.db_file) as con:
con.row_factory = sqlite3.Row
cur = con.cursor()
con.text_factory = str
if expire_stale:
self.expire_stale_records(cur, "Series")
# fetch
cur.execute("SELECT * FROM Series WHERE id=? AND source=?", [series_id, source])
row = cur.fetchone()
if row is None:
return None
result = Series(id=row["id"], data=row["data"])
return (result, row["complete"])
def get_series_issues_info(
self, series_id: str, source: str, expire_stale: bool = True
) -> list[tuple[Issue, bool]]:
with sqlite3.connect(self.db_file) as con:
con.row_factory = sqlite3.Row
cur = con.cursor()
con.text_factory = str
if expire_stale:
self.expire_stale_records(cur, "Issues")
# fetch
results: list[tuple[Issue, bool]] = []
cur.execute("SELECT * FROM Issues WHERE series_id=? AND source=?", [series_id, source])
rows = cur.fetchall()
# now process the results
for row in rows:
record = (Issue(id=row["id"], series_id=row["series_id"], data=row["data"]), row["complete"])
results.append(record)
return results
def get_issue_info(self, issue_id: int, source: str, expire_stale: bool = True) -> tuple[Issue, bool] | None:
with sqlite3.connect(self.db_file) as con:
con.row_factory = sqlite3.Row
cur = con.cursor()
con.text_factory = str
if expire_stale:
self.expire_stale_records(cur, "Issues")
cur.execute("SELECT * FROM Issues WHERE id=? AND source=?", [issue_id, source])
row = cur.fetchone()
record = None
if row:
record = (Issue(id=row["id"], series_id=row["series_id"], data=row["data"]), row["complete"])
return record
def upsert(self, cur: sqlite3.Cursor, tablename: str, data: dict[str, Any]) -> None:
"""This does an insert if the given PK doesn't exist, and an
update it if does
TODO: should the cursor be created here, and not up the stack?
"""
keys = ""
vals = []
ins_slots = ""
set_slots = ""
for key in data:
if data[key] is None:
continue
if keys != "":
keys += ", "
if ins_slots != "":
ins_slots += ", "
if set_slots != "":
set_slots += ", "
keys += key
vals.append(data[key])
ins_slots += "?"
set_slots += key + " = ?"
sql_ins = f"INSERT OR REPLACE INTO {tablename} ({keys}) VALUES ({ins_slots})"
cur.execute(sql_ins, vals)