import type { DatabaseSync } from "node:sqlite";

// Repairs 0011, which was too eager about what counts as a site icon.
//
// 0011 tested for "favicon" anywhere in the path, so a legitimate `og:image`
// like mirtitles.org's `/wp-content/uploads/2022/04/mir-logo-favicon.png` was
// moved into `faviconUrl` and its `thumbnailUrl` cleared — which also demoted
// the dump from an image card to a pull-quote in the journal mosaic. A blanket
// `.svg` match had the same effect on real artwork.
//
// `isIconUrl` here looks at the *filename* rather than the whole path: an icon
// is named `favicon…`, `apple-touch-icon…` or `icon…`, lives in an `icons/`
// directory, or ends in `.ico`. That both spares `mir-logo-favicon.png` and
// catches `icon_SEARCH.png`, which 0011 missed.
//
// Applied in both directions, and deliberately narrow when reversing: a value
// is only promoted back to `thumbnailUrl` if it matches 0011's rule but not
// this one, so rows written by a normal fetch aren't disturbed.

/** 0011's rule, kept verbatim so its false positives can be identified. */
function wasIconUrl0011(raw: string): boolean {
  let pathname: string;
  try {
    pathname = new URL(raw).pathname;
  } catch {
    return false;
  }
  return /favicon|apple-touch-icon|\/icons?\//i.test(pathname) ||
    /\.(ico|svg)$/i.test(pathname);
}

/** Whether a URL names a site icon, judged on its filename and directory. */
function isIconUrl(raw: string): boolean {
  let pathname: string;
  try {
    pathname = new URL(raw).pathname;
  } catch {
    return false;
  }
  const segments = pathname.split("/").filter(Boolean);
  const filename = segments.pop() ?? "";
  if (/\.ico$/i.test(filename)) return true;
  if (segments.some((s) => /^(favicons?|icons?)$/i.test(s))) return true;
  const basename = filename.replace(/\.[^.]+$/, "");
  return /^(favicon|apple-touch-icon|icon)(?=$|[-_.\d])/i.test(basename);
}

export function up(db: DatabaseSync): void {
  const rows = db.prepare(
    `SELECT id, rich_content FROM dumps
      WHERE kind = 'url' AND rich_content IS NOT NULL;`,
  ).all() as { id: string; rich_content: string }[];

  const update = db.prepare(
    `UPDATE dumps SET rich_content = ? WHERE id = ?;`,
  );

  let restored = 0;
  let reclassified = 0;
  for (const row of rows) {
    let rich: { thumbnailUrl?: string; faviconUrl?: string };
    try {
      rich = JSON.parse(row.rich_content);
    } catch {
      continue; // malformed payload — leave it untouched
    }

    // Artwork 0011 mistook for an icon: put it back.
    if (
      rich.faviconUrl && !rich.thumbnailUrl &&
      wasIconUrl0011(rich.faviconUrl) && !isIconUrl(rich.faviconUrl)
    ) {
      const { faviconUrl: _dropped, ...rest } = rich;
      update.run(
        JSON.stringify({ ...rest, thumbnailUrl: rich.faviconUrl }),
        row.id,
      );
      restored++;
      continue;
    }

    // An icon 0011's rule didn't recognize: move it now.
    if (
      rich.thumbnailUrl && !rich.faviconUrl && isIconUrl(rich.thumbnailUrl)
    ) {
      const { thumbnailUrl: _dropped, ...rest } = rich;
      update.run(
        JSON.stringify({ ...rest, faviconUrl: rich.thumbnailUrl }),
        row.id,
      );
      reclassified++;
    }
  }

  if (restored > 0 || reclassified > 0) {
    console.log(
      `[migrate] 0012: restored ${restored} thumbnail(s), reclassified ${reclassified} icon(s)`,
    );
  }
}