bibdle/src/lib/server/rss.ts

import type { DailyVerse } from '$lib/server/db/schema';

/**
 * The feed presents calendar days, so every timestamp is anchored to midnight
 * in this reference timezone. Using a fixed zone (rather than the server's
 * local timezone) keeps output stable across UTC dev machines and NY hosts.
 */
export const FEED_TIMEZONE = 'America/New_York';

/** Number of most-recent verses exposed in the feed. */
export const FEED_ITEM_LIMIT = 30;

/** Escape the five XML predefined entities. */
export function escapeXml(text: string): string {
	return text
		.replace(/&/g, '&')
		.replace(/</g, '&lt;')
		.replace(/>/g, '&gt;')
		.replace(/"/g, '&quot;')
		.replace(/'/g, '&apos;');
}

/**
 * The current calendar date (`YYYY-MM-DD`) in the feed's reference timezone.
 * Used to cap the feed so future/scheduled verses are never published early.
 */
export function todayInTimeZone(timeZone: string = FEED_TIMEZONE, now: Date = new Date()): string {
	const parts = new Intl.DateTimeFormat('en-US', {
		timeZone,
		year: 'numeric',
		month: '2-digit',
		day: '2-digit'
	}).formatToParts(now);
	const get = (type: string) => parts.find((part) => part.type === type)?.value ?? '';
	return `${get('year')}-${get('month')}-${get('day')}`;
}

/**
 * The UTC offset in effect on a given `YYYY-MM-DD` for a timezone, formatted
 * as `±HH:MM` (e.g. `-04:00` during EDT, `-05:00` during EST).
 */
export function timeZoneOffset(dateStr: string, timeZone: string = FEED_TIMEZONE): string {
	const [year, month, day] = dateStr.split('-').map(Number);
	// Noon UTC is safely inside the calendar day in any US offset, so DST is
	// resolved correctly regardless of the time of day.
	const reference = new Date(Date.UTC(year, month - 1, day, 12));
	const name = new Intl.DateTimeFormat('en-US', {
		timeZone,
		timeZoneName: 'longOffset'
	})
		.formatToParts(reference)
		.find((part) => part.type === 'timeZoneName')?.value;

	const offset = (name ?? 'GMT+00:00').replace('GMT', '');
	return offset === '' ? '+00:00' : offset;
}

/**
 * Format a `YYYY-MM-DD` day as an RFC 822 / RFC 2822 timestamp at midnight in
 * the feed's reference timezone. Always emits a valid numeric/GMT timestamp
 * (never a bare, ambiguous abbreviation like `EST`).
 */
export function formatRFC822(dateStr: string, timeZone: string = FEED_TIMEZONE): string {
	const offset = timeZoneOffset(dateStr, timeZone);
	return new Date(`${dateStr}T00:00:00${offset}`).toUTCString();
}

/** Human-readable date that does not shift with the server's local timezone. */
export function formatReadableDate(dateStr: string): string {
	const [year, month, day] = dateStr.split('-').map(Number);
	return new Date(Date.UTC(year, month - 1, day)).toLocaleDateString('en-US', {
		month: 'long',
		day: 'numeric',
		year: 'numeric',
		timeZone: 'UTC'
	});
}

/** Wrap text in a CDATA section, splitting any embedded `]]>` safely. */
export function cdata(text: string): string {
	return `<![CDATA[${text.replace(/]]>/g, ']]]]><![CDATA[>')}]]>`;
}

// Helper: Format verse text (VerseDisplay + Imposter unbalanced punctuation handling)
export function formatVerseText(text: string): string {
	let formatted = text;

	// Handle unbalanced opening/closing punctuation (from Imposter.svelte)
	const pairs: [string, string][] = [
		['(', ')'],
		['[', ']'],
		['{', '}'],
		['"', '"'],
		["'", "'"],
		['\u201C', '\u201D'], // " "
		['\u2018', '\u2019'] // ' '
	];

	// Check if text starts with opening punctuation without closing
	for (const [open, close] of pairs) {
		if (formatted.startsWith(open) && !formatted.includes(close)) {
			formatted += '...' + close;
			break;
		}
	}

	// Check if text ends with closing punctuation without opening
	for (const [open, close] of pairs) {
		if (formatted.endsWith(close) && !formatted.includes(open)) {
			formatted = open + '...' + formatted;
			break;
		}
	}

	// Check if text contains unbalanced opening quotes (not at start) without closing
	for (const [open, close] of pairs) {
		const openCount = (formatted.match(new RegExp(`\\${open}`, 'g')) || []).length;
		const closeCount = (formatted.match(new RegExp(`\\${close}`, 'g')) || []).length;
		if (openCount > closeCount) {
			formatted += close;
			break;
		}
	}

	// Capitalize first letter if lowercase (from VerseDisplay.svelte)
	formatted = formatted.replace(/^([a-z])/, (c) => c.toUpperCase());

	// Replace trailing punctuation with ellipsis
	// Preserve closing quotes/brackets that may have been added
	formatted = formatted.replace(/[,:;-—]([)\]}\"\'\u201D\u2019]*)$/, '...$1');

	return formatted;
}

/**
 * Content fingerprint used as a strong ETag. Any change to the included rows
 * (new day, edited verse text, changed reference) produces a new hash, so
 * cached clients revalidate correctly.
 */
export function feedFingerprint(
	verses: Pick<DailyVerse, 'date' | 'reference' | 'verseText'>[]
): string {
	let hash = 5381;
	for (const verse of verses) {
		const row = `${verse.date}\u0000${verse.reference}\u0000${verse.verseText}`;
		for (let i = 0; i < row.length; i++) {
			hash = ((hash << 5) + hash + row.charCodeAt(i)) | 0;
		}
	}
	return (hash >>> 0).toString(16);
}

/**
 * The most recent `createdAt` across the given rows, used for `Last-Modified`.
 * Falls back to `fallback` when no row carries a timestamp.
 */
export function latestModification(
	verses: Pick<DailyVerse, 'createdAt'>[],
	fallback: Date
): Date {
	return verses.reduce<Date | null>((latest, verse) => {
		if (!verse.createdAt) return latest;
		return !latest || verse.createdAt > latest ? verse.createdAt : latest;
	}, null) ?? fallback;
}