Monocle Editorial Search

Purpose

Search the Monocle editorial archive (monocle.com) for articles matching a query — title, canonical URL, author byline, publication date, primary topic, category tags, excerpt, and (optionally) full article body. Optionally filter by topic (Affairs, Design, Travel, ...) and exclude non-editorial formats (radio episodes, city guides, events, partnered content). Read-only — never logs in, never modifies state. Copenhagen is the canonical example query; the skill generalises to any city, place, person, or keyword Monocle has written about.

When to Use

• "What has Monocle written about Copenhagen?" / "Find Monocle's design coverage of Tokyo." / "List recent Monocle articles tagged urbanism."
• Building a research dossier of Monocle's editorial coverage of a city or topic.
• Periodic monitoring of new Monocle editorials on a watch-term (combine with pubDate from RSS to detect new items since last poll).
• Bulk extraction across many query terms — RSS path is cheap (~150KB per page, 10 items, plain HTTP fetch, no auth, no anti-bot).

Workflow

Monocle is a public WordPress site (Automattic VIP — X-Hacker header) with ElasticPress-backed search (X-Elasticpress-Query: true on responses). The official WP REST API is disabled (/wp-json/wp/v2/posts?search=... → 404 rest_no_route, despite the Link: <https://monocle.com/wp-json/>; rel="https://api.w.org/" header advertising it). However, the per-query RSS feed is enabled and returns richer data than the HTML search page — most notably it includes the <dc:creator> author byline and <content:encoded> full-body HTML, both of which are absent from the HTML article-card markup. Lead with the RSS path; HTML browse is a fallback when you also need featured-image URLs or the total result count.

There is no anti-bot wall: bare browse cloud fetch (no --proxies, no --verified) returns 200 OK from both the HTML and RSS endpoints. Cookie consent is JS-only and never blocks the underlying HTML/XML body.

Recommended: RSS feed (per-query)

1. Build the query URL. Two interchangeable shapes both work:

   • Query string: https://monocle.com/feed/?s={URL-enc query}&search_format=post[&search_topic={slug}][&paged={N}]

   • Path style: https://monocle.com/search/{URL-enc query}/feed/?search_format=post[&search_topic={slug}][&paged={N}]

   • s (or path segment): the search term.

   • search_format=post: the editorial filter — restricts to WordPress posts (i.e. magazine articles), excluding event, travel_guide, radio_episode, partnered_content. Omit this param to return all formats.

   • search_topic={slug}: optional single-topic facet (e.g. design, affairs, urbanism, travel-and-restaurants). See the topic-slug list in "Site-Specific Gotchas".

   • paged=N: 1-indexed page. Each page returns 10 <item> blocks. Walking past the last page returns HTTP 404 — a clean termination signal.

2. Fetch:

   browse cloud fetch "https://monocle.com/feed/?s=copenhagen&search_format=post&paged=1"
   

   No --proxies, no session, no cookies needed. Response is application/rss+xml; charset=UTF-8, ~120-150 KB per page for 10 items including full bodies.

3. Parse each <item>:

   • <title> — article title (HTML-entity decode required: e.g. ’ → ').
   • <link> — canonical article URL (https://monocle.com/{topic}/{slug}/).
   • <dc:creator> — author byline (CDATA-wrapped; RSS-only, not in HTML cards).
   • <pubDate> — RFC-2822 timestamp (e.g. Fri, 20 Jun 2025 18:29:50 +0000).
   • <category> (repeated 1-N times) — primary topic comes first, followed by tag slugs. First category is the same value rendered as the topic badge in the HTML.
   • <description> — CDATA-wrapped HTML excerpt (1-2 sentences). Strip the trailing The post <a>...</a> appeared first on... boilerplate.
   • <content:encoded> — CDATA-wrapped full article body HTML. Use only if you need the body; otherwise skip — it's ~10-15 KB per item.

4. Paginate until HTTP 404 is returned by paged=N. Result count is not exposed in RSS — if you need the total up-front, hit the HTML page once (step below) and parse the count selector before walking RSS.

Browser fallback: HTML search page

Use when you need featured-image URLs (not in RSS) or the up-front total-results count, or when the RSS feed is unreachable.

1. Build the URL (same param surface as RSS, no /feed/ segment):

   https://monocle.com/?s={URL-enc query}&search_format=post[&search_topic={slug}][&paged={N}]
   

   Or path style: https://monocle.com/search/{query}[/page/{N}/][?search_format=post].

2. Fetch with browse cloud fetch <url> — no stealth needed. Or drive interactively with browse open <url> if you want screenshots/snapshots for debugging.

3. Parse the HTML:

   • Total count: <div class="o-search-results__actions"> <p>{N} stories about "{query}"</p> → regex (\d+)\s+stories about\s+["“]([^"”]+)["”].
   • Each result card: <article id="{POST_ID}" class="c-article-card ...">. The id attribute is the stable WordPress post ID — use it for deduping.
   • Within each card:
     • Category badge: span.c-article-card__category a — href is the topic URL, text is the topic name.
     • Title + URL: h3.c-article-card__title a — href is the canonical article URL, text is the title.
     • Excerpt: p.c-article-card__description.
     • Meta items: ul.c-article-card__meta li — each <li> may begin with an inline SVG decoration; strip inner tags before reading text (e.g. Issue #185, 3 min read). Naive <li>([^<]+)</li> regex skips Issue-# items because of the leading SVG.
     • Featured image: figure.c-article-card__image img — src and srcset (1x / 2x).
   • Pagination: nav block with class posts-pagination; next page is https://monocle.com/search/{query}/page/{N+1}/ (preserves any ?search_format / ?search_topic query params).

Site-Specific Gotchas

• WP REST API is disabled despite advertising itself. Every /wp-json/wp/v2/* route returns {"code":"rest_no_route","status":404}, even though the response headers include X-WP-Total, X-WP-TotalPages, Access-Control-Allow-Headers: X-WP-Nonce, and a Link header pointing to /wp-json/. Don't waste cycles probing alternate REST routes — the site has stripped them at the WordPress level. Use the RSS feed instead.
• search_topic[] array notation is silently ignored. ?s=copenhagen&search_topic[]=design&search_topic[]=culture returns the unfiltered set (171 results), not the union (the design-only subset is 49). Only single-value search_topic=<slug> filtering works through the URL layer. To collect across multiple topics, issue separate requests per topic and dedupe by post ID (<article id="...").
• The Apply-Filters button in the UI drops the search query. Clicking the FILTER button on a search-results page, selecting a format, and pressing APPLY FILTERS navigates to https://monocle.com/?search_format=post — the s={query} param is discarded. Always build URLs directly with both params rather than relying on the in-page filter UI.
• "Editorials" = search_format=post. Monocle's UI calls them "Article" but the underlying WP post-type slug is post. The other four format slugs (event, travel_guide, radio_episode, partnered_content) are not editorial content and should be excluded for an editorials-only query. Omitting search_format returns the union of all five.
• Author bylines are in RSS only. The HTML article-card markup (.c-article-card) has no author element. If you need the byline, you must hit the RSS feed (or click through to the individual article page).
• Featured image URLs are in HTML only. The RSS feed has no <media:content> or <enclosure> elements. If you need thumbnails, scrape figure.c-article-card__image img from the HTML page.
• Per-page size is fixed at 10. Both HTML pagination (/page/N/) and RSS pagination (?paged=N) return 10 items per page. There is no per-page override (per_page=, posts_per_page=, etc.).
• Pagination past the last page returns HTTP 404 for RSS and a rendered "no results" HTML page for the search route. Use 404 (RSS) or the absence of .c-article-card blocks (HTML) as the loop-termination signal.
• Issue-# meta items contain a leading inline SVG. Inside ul.c-article-card__meta, items like <li><svg>...</svg> Issue #185 </li> will be missed by a <li>([^<]+)</li> regex. Either parse as DOM and read textContent, or use a regex that strips inner <svg>…</svg> first. Read-time items (3 min read) have no leading SVG and parse cleanly.
• HTML entities in titles. RSS-feed titles are entity-encoded (Copenhagen’s for Copenhagen's). Decode before emitting.
• description carries boilerplate. The RSS <description> ends with <p>The post <a>...</a> appeared first on <a href="https://monocle.com">Monocle</a>.</p> — strip this paragraph for a clean excerpt.
• content:encoded is large. Each item's full-body HTML is ~10-15 KB. If you only need title + URL + date, parse only the elements you need rather than the full item. For bulk runs, prefer reading the RSS feed once and persisting parsed items rather than re-fetching.
• Format slugs (search_format): post (Article — editorial), event, travel_guide (City Guide), radio_episode, partnered_content.
• Topic slugs (search_topic, observed from the filter modal's data-value attributes): affairs, architecture, art, arts, aviation, books, business, craft, culture, defence, design, diplomacy, economics, economy, education, entertaining, entertainment, entrepreneurialism, environment, fashion, film, food-drink, furniture, government, health, hospitality, industry, konfekt, manufacturing, media, monocle-films, monocle-radio, music, photography, politics, product-design, property, recipe, residences, retail, shoots, society, soft-power, sport, technology, the-faster-lane, the-monocle-concierge, the-monocle-minute, the-weekend-opener, transport, travel-and-restaurants, urbanism, wine. (The label shown in the filter UI is the title-cased slug with hyphens replaced by spaces.)
• ?s= vs /search/{query} are equivalent. Both forms hit the same handler and produce identical results. Path-style URLs are slightly cleaner for direct linking; query-style is easier to build programmatically.
• No geo-redirect, no IP scoping, no rate-limit observed in test. Run from anywhere; keep ≤ 1 req/s sustained as a courtesy.

Expected Output

{
  "query": "copenhagen",
  "format": "post",
  "topic": null,
  "total_results": 171,
  "page": 1,
  "items": [
    {
      "post_id": 195123,
      "title": "Why Copenhagen's 3 Days of Design leaves such a lasting impression",
      "url": "https://monocle.com/design/3-days-of-design-copenhagen-comment/",
      "author": "Kate Lucey",
      "published_at": "2025-06-20T18:29:50Z",
      "primary_topic": "Design",
      "categories": ["Design", "3 days of design", "design fairs"],
      "excerpt": "Designers from Tokyo to Porto headed to Copenhagen to rethink what a design fair can be, with thoughtful collaborations and intimate, idea-led showcases.",
      "issue": null,
      "read_time_minutes": null,
      "image_url": "https://monocle.com/wp-content/uploads/2025/06/EIS_20250617_1313_CROP.jpg?w=745"
    },
    {
      "post_id": 189311,
      "title": "Copenhagen's latest park demonstrates the virtues of having no kids on the block",
      "url": "https://monocle.com/affairs/urbanism/copenhagens-adult-only-opera-park/",
      "author": "Carlota Rebelo",
      "published_at": "2025-06-15T09:00:00Z",
      "primary_topic": "Urbanism",
      "categories": ["Urbanism", "parks", "denmark"],
      "excerpt": "Inside the sanctuary of Opera Park, a child-free green space designed strictly for grown-ups.",
      "issue": "185",
      "read_time_minutes": 3,
      "image_url": "https://monocle.com/wp-content/uploads/2025/06/Monocle_Skip_Final_LargerBG_thumb.jpg?w=745"
    }
  ],
  "next_page": "https://monocle.com/feed/?s=copenhagen&search_format=post&paged=2"
}

Outcome shapes:

// No results for the query
{ "query": "asdfqwerzxcv", "format": "post", "total_results": 0, "items": [] }

// Past last page (RSS 404)
{ "query": "copenhagen", "format": "post", "page": 99, "items": [], "end_of_results": true }

// Topic filter applied
{ "query": "copenhagen", "format": "post", "topic": "design", "total_results": 49, "items": [...] }

// All formats (omit search_format)
{ "query": "copenhagen", "format": null, "total_results": 352, "items": [...] }

Notes on the JSON above: issue and read_time_minutes come from the HTML ul.c-article-card__meta block and are null on items not tied to a print issue (e.g. web-only comment pieces — id=195123 above is one). image_url is HTML-only; pure-RSS callers will see image_url: null. author is RSS-only; pure-HTML callers will see author: null. For a complete record, run the RSS feed and HTML page once each and merge on post_id (the <article id> attribute on HTML matches the WP post ID; RSS items don't expose the ID directly — match by canonical URL slug).