a\n const titleMatch = fullArticle.match(/

([\\s\\S]?)<\\/a><\\/h5>/);\n if (!titleMatch) continue;\n\n const url = titleMatch[1];\n const title = decodeHtmlEntities(stripHtml(titleMatch[2]));\n\n // Extract time\n const timeMatch = fullArticle.match(\n /([\\s\\S]?)<\\/span><\\/strong>/\n );\n let time = \"\";\n let endTime: string | null = null;\n if (timeMatch) {\n const timeHtml = timeMatch[1];\n // Start time is the text before any separator\n const startMatch = timeHtml.match(/^([^<]+)/);\n time = startMatch ? startMatch[1].trim() : \"\";\n // End time is in span.end-time\n const endMatch = timeHtml.match(/(.?)<\\/span>/);\n endTime = endMatch ? endMatch[1].trim() : null;\n }\n\n // Extract badges/categories from span.categories\n const badgesMatch = fullArticle.match(/([\\s\\S]?)<\\/span>
/);\n const badges: string[] = [];\n if (badgesMatch) {\n const badgeText = badgesMatch[1];\n // Badge names appear after tags\n const badgeNames = badgeText.match(/<\\/title>([^<,]+)/g);\n if (badgeNames) {\n for (const b of badgeNames) {\n const name = b.replace(\"\", \"\").trim();\n if (name) badges.push(name);\n }\n }\n }\n\n // Extract description from toggle content\n const descMatch = fullArticle.match(\n /
([\\s\\S]?)<\\/div>\\s<\\/div>/\n );\n let description = \"\";\n if (descMatch) {\n // Get everything between the toggle content divs, skip the first
(which is just the badge repeat)\n const descHtml = descMatch[1];\n // Find the second
block which has the actual description\n const pBlocks = descHtml.split(/<\\/p>\\s*
/);\n if (pBlocks.length >= 2) {\n // The description is in the second p block (index 1)\n description = decodeHtmlEntities(stripHtml(pBlocks[1]));\n } else {\n description = decodeHtmlEntities(stripHtml(descHtml));\n }\n }\n\n // Clean up description - remove \"Add this event to your calendar\" and after\n const calIdx = description.indexOf(\"Add this event to your calendar\");\n if (calIdx > 0) description = description.slice(0, calIdx).trim();\n\n // Remove [email protected] artifacts\n description = description.replace(/\\[email\\sprotected\\]/g, \"[email]\");\n\n events.push({\n title,\n time,\n endTime,\n url,\n description,\n badges,\n date: dateStr,\n });\n }\n\n return events;\n}\n\n// ── Scraping ───────────────────────────────────────────────────────────\n\nexport function formatDate(d: Date): { dateStr: string; dy: string; month: string; yr: string } {\n const yr = d.getFullYear().toString();\n const month = (d.getMonth() + 1).toString().padStart(2, \"0\");\n const dy = d.getDate().toString();\n const dateStr = `${yr}-${month}-${dy.padStart(2, \"0\")}`;\n return { dateStr, dy, month, yr };\n}\n\nasync function scrapeDay(date: Date): Promise {\n const { dateStr, dy, month, yr } = formatDate(date);\n const url = `https://annarborobserver.com/calendar/?cid=main-calendar&dy=${dy}&mcat=all&month=${month}&time=day&yr=${yr}`;\n\n console.log(`Fetching ${dateStr}...`);\n\n const resp = await fetch(url);\n if (!resp.ok) {\n console.error(`Failed to fetch ${dateStr}: ${resp.status}`);\n return [];\n }\n const html = await resp.text();\n const events = parseEventsFromHtml(html, dateStr);\n console.log(` → parsed ${events.length} events for ${dateStr}`);\n return events;\n}\n\nexport async function scrapeAllDays(): Promise {\n const allEvents: RawEvent[] = [];\n const today = new Date();\n\n for (let i = 0; i < 8; i++) {\n const d = new Date(today);\n d.setDate(d.getDate() + i);\n const events = await scrapeDay(d);\n allEvents.push(...events);\n // Be polite: small delay between requests\n if (i < 7) await new Promise((r) => setTimeout(r, 500));\n }\n\n return allEvents;\n}\n\n// ── SQLite storage ─────────────────────────────────────────────────────\n\nexport async function initDb(dropFirst = false) {\n if (dropFirst) {\n await sqlite.execute(`DROP TABLE IF EXISTS events`);\n }\n await sqlite.execute(`\n CREATE TABLE IF NOT EXISTS events (\n id INTEGER PRIMARY KEY AUTOINCREMENT,\n title TEXT,\n time TEXT,\n end_time TEXT,\n url TEXT,\n description TEXT,\n badges TEXT,\n date TEXT,\n category TEXT,\n scraped_at TEXT\n )\n `);\n await sqlite.execute(`\n CREATE INDEX IF NOT EXISTS idx_events_date ON events(date)\n `);\n}\n\nexport async function storeEvents(events: ClassifiedEvent[]) {\n // Clear existing events for the dates we're updating\n const dates = [...new Set(events.map((e) => e.date))];\n for (const date of dates) {\n await sqlite.execute({ sql: `DELETE FROM events WHERE date = ?`, args: [date] });\n }\n\n const now = new Date().toISOString();\n let stored = 0;\n for (const e of events) {\n if (!e.title || String(e.title).trim() === \"\") {\n console.warn(` [skip] Event with empty title on ${e.date}`);\n continue;\n }\n try {\n await sqlite.execute({\n sql: `INSERT INTO events (title, time, end_time, url, description, badges, date, category, scraped_at)\n VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?)`,\n args: [String(e.title), e.time || \"\", e.endTime || null, e.url || \"\", e.description || \"\", JSON.stringify(e.badges || []), e.date, e.category || \"other\", now],\n });\n stored++;\n } catch (err) {\n console.error(` [error] Failed to store \"${e.title}\":`, err);\n }\n }\n\n console.log(`Stored ${stored}/${events.length} events across ${dates.length} days`);\n}\n\nexport async function getEvents(from: string, to: string) {\n const result = await sqlite.execute({\n sql: `SELECT FROM events WHERE date >= ? AND date <= ? ORDER BY date, time`,\n args: [from, to],\n });\n\n return result.rows.map((row: any) => ({\n id: row[0],\n title: row[1],\n time: row[2],\n endTime: row[3],\n url: row[4],\n description: row[5],\n badges: JSON.parse(row[6] || \"[]\"),\n date: row[7],\n category: row[8],\n scrapedAt: row[9],\n }));\n}\n","deleted","label",[100],"permalink","https://www.val.town/x/dm0n3y/a2-observer/code/lib.ts","extension","actionData","errors"]