2nd STREET

Search 2nd STREET. ONE ROW IS ONE PHYSICAL SECOND-HAND PIECE held by ONE NAMED SHOP — the chain buys outright and resells, so there is no seller, no variant and no stock count; the size IS the row and every Japanese row carries the `shop_id` that holds it. JAPAN: 60 rows a page, fixed (the source swallowed per_page/limit/count and served 60 when 120 was asked for), and the WHOLE result set is walkable — page 20,000 of the unfiltered catalogue still returned 60 real rows and the page after the last one returns an EMPTY grid, never a repeat. Prices are whole yen and TAX-INCLUDED: `税込` appears on every product page measured and `税抜` on none, and the grid price was checked against 12 product pages' own microdata (12/12 identical). Japanese text — size, condition, colour, material — is republished exactly as the source wrote it; nothing is translated. UNITED STATES: the storefront's search answers at most 1,000 results (it prints '1000 results found' for nike, jacket and even `*`) over 28 pages of 36, so a large query ends with `stop_reason: cap_reached`. Pass `collection` instead of `keyword` to walk a collection without that cap, up to 250 rows a page; that surface additionally publishes `tags` (which carry the US branch name), `images` and `compare_at_price_usd`, which the keyword surface does not print.

One piece in full. JAPAN publishes, and this returns separately: the printed tax-inclusive price, the pre-markdown price (通常価格) with the discount percentage the page prints, and the shipping charge — either the flat ¥770 (tax included, its own label) or the marker that the piece carries 個別送料, an individual delivery charge, in which case `shipping_jpy` is null rather than guessed. The condition comes back three ways with no translation invented: `condition` is the source's own text (中古B), `condition_grade` is the source's OWN filter code (B), and `condition_note` is the sentence the page prints about that grade. Attributes (colour, pattern, material, model number, size) are parsed out of the page's own 商品の詳細 table, which is category-shaped — a laptop prints OS/CPU/memory where a jacket prints 柄 and 素材・生地 — so the whole table is also returned verbatim as `specs`, plus the tailor's measurements in cm as `measurements_cm` (a figure printed as `/` means not measured and is omitted, not 0). 🔴 THE TWO STOCK SURFACES ARE KEPT APART: `available_online` is whether the page offers a cart, `store_transfer` is whether it offers 店舗取り寄せ to another branch, and `shop` is the branch that physically holds it, with its address, opening hours, buy-back hours and parking. A SOLD piece does not go OutOfStock — its id starts serving the shop's landing page with HTTP 200, and that is returned as NOT_FOUND here rather than as a success carrying somebody else's merchandise. UNITED STATES: price in dollars and in cents (the store publishes the same number both ways — 24990 and "249.90" — and they are never mixed), availability, the branch name from the product's own tags, and the Size/Color/Material/Condition table the store writes into the description.

The shop's own category tree with its own item counts. JAPAN: 16 top-level departments over three levels, every node carrying the count the chain publishes for it — メンズウェア 680,072 · シューズ 256,471 · レディースウェア 199,572 · 服飾雑貨他 178,470 · バッグ 139,302 · 家電・ビジュアル・オーディオ 40,736 · スポーツ 32,666 · ホビー 27,098 · キッチン用品 20,184 · 楽器 6,828 · カメラ 6,413 · インテリア小物 6,730 · キッズ 6,244 · パソコン 5,975 · DIY用品 1,676 · 家具 1,071, and a catalogue total of 1,609,533 (2026-10-08). Names are Japanese because that is what the source publishes; no translation is invented. The same call also returns the 47 prefectures, the 11 sales areas and the 23-colour palette the search filters use. UNITED STATES: the store's collections with their handles — pass one back as `search(collection=…)`. 🔴 Their own `products_count` is republished as the source prints it and is NOT a measurement: the biggest collection claims 2,357,882 products on a store whose search surface caps at 1,000 and whose whole product list was walked to 15,000, so the figure is a Shopify smart-collection count, not a catalogue size.

2nd STREET Japan's whole brand dictionary — 14,980 brands, each with BOTH spellings the chain itself publishes (`brand` in Latin script, `brand_kana` in Japanese katakana; 14,980 of 14,980 rows carry both, measured) plus the id the `search(brand=…)` filter takes and the live item count. Biggest measured: NIKE/ナイキ 101,313 · THE NORTH FACE/ザノースフェイス 54,088 · LOUIS VUITTON/ルイヴィトン 35,509 · adidas/アディダス 26,169 · Levi's/リーバイス 24,333 · Supreme/シュプリーム 23,718. 4,180 brands currently show a count of 0 — they are in the dictionary with nothing in stock, which is an answer, so `min_count` is offered rather than silently dropping them. With `keyword` this is also the chain's own brand look-up: it matches the katakana reading, so ナイキ returns NIKE, NIKE ACG and NIKE SB with their counts. 🔴 The id 999999 (その他ブランド, 119,435 items) is a real, usable filter value that this dictionary does not list. JAPAN ONLY — the US store publishes no brand index.

What the chain's own search index says about a query BEFORE you page through it: the total, and the breakdown by brand, category, colour, condition grade and size, each with its own count. This is the same figure the result page prints (103,034 vs 103,036 on the grid, a live-index difference of 2) and it is the cheap way to size a query, measure a filter or build a facet UI without pulling 60-row pages. Takes exactly the same filters as `search`, so it also answers 'how much would this filter bite'. The colour buckets come back with the shop's own Japanese colour names attached, the condition buckets with the source's own 中古A/中古B/… labels, and the size buckets carry the label the index itself publishes (`XXS`, `L`, `28cm`); brand and category buckets are ids, resolvable with `brands` and `categories`. JAPAN ONLY.

2nd STREET Japan's shop network. With no argument this returns the chain's own shop list — 985 entries with their ids and house names, against a headline figure of 971 shops that the chain publishes itself (the list carries a few non-trading placeholders, so both numbers are returned rather than one being quietly preferred) — together with the 47 prefectures, the 11 sales areas, the cities of each prefecture, and how many duty-free shops each prefecture has. With `prefecture` it returns that prefecture's shops only, from the chain's own per-prefecture endpoint: Tokyo (13) answers with 70 shops. Feed a `shop_id` straight back into `search(shop=…)` — it bites (30251 → 2,296 items) — or into `store` for that branch's full record. JAPAN ONLY: the US site's shop finder sits behind a JavaScript interstitial and publishes no catalogue.

One 2nd STREET branch in full, as its own page prints it: house name and shop name, shop type (トータルリユース, 買取専門店 …), whether it is a duty-free shop, its description, street address, telephone number, opening hours, buy-back reception hours, visiting buy-back hours where it offers them, parking spaces, the latitude and longitude from the map the page embeds, its area and prefecture, its photographs, and the two lists of goods it handles — what it SELLS and what it BUYS BACK, which are not the same list. Pair the `shop_id` with `search(shop=…)` to see what that branch has on sale right now. JAPAN ONLY.

The US store's own type-ahead: the products, the collections and the alternative search phrases it would offer a shopper mid-keystroke, each product already carrying its price, availability, image and the Size/Color/Material/Condition table the store writes into its description. Capped by the store at 10 results per kind (20 was asked for and 10 came back). UNITED STATES ONLY — the Japanese site has no suggest endpoint (its /searchapi/suggest answers 200 with an empty body); for Japan, `brands(keyword=…)` is the equivalent look-up and it matches the katakana reading.