storage.py

\
import sqlite3, time, os

SCHEMA = """
CREATE TABLE IF NOT EXISTS items(
  id INTEGER PRIMARY KEY,
  fingerprint TEXT UNIQUE,
  source TEXT,
  title TEXT,
  url TEXT,
  published_at INTEGER,
  symbols TEXT,
  tags TEXT
);
CREATE INDEX IF NOT EXISTS idx_published ON items(published_at);
"""

class Storage:
    def __init__(self, path: str):
        self.conn = sqlite3.connect(path)
        self.conn.execute("PRAGMA journal_mode=WAL;")
        self.conn.executescript(SCHEMA)

    def seen(self, fingerprint: str) -> bool:
        cur = self.conn.execute("SELECT 1 FROM items WHERE fingerprint=?", (fingerprint,))
        return cur.fetchone() is not None

    def add(self, fingerprint: str, source: str, title: str, url: str, published_at: int, symbols: str, tags: str):
        self.conn.execute(
            "INSERT OR IGNORE INTO items(fingerprint, source, title, url, published_at, symbols, tags) VALUES(?,?,?,?,?,?,?)",
            (fingerprint, source, title, url, published_at, symbols, tags),
        )
        self.conn.commit()

    def purge_old(self, keep_seconds: int):
        cutoff = int(time.time()) - keep_seconds
        self.conn.execute("DELETE FROM items WHERE published_at < ?", (cutoff,))
        self.conn.commit()