| name | livepocket |
| description | Platform rules, search URL structure, dl field extraction, and Taiwan filter for the LivePocket（ライブポケット）e-ticket scraper |
| applyTo | scraper/sources/livepocket.py |
LivePocket（ライブポケット）Scraper
Platform Profile
| Field | Value |
|---|
| Site URL | https://livepocket.jp/ |
| Search URL | /event/search?word=台湾&page=N |
| Rendering | Static HTML — no JavaScript required |
| Auth | None required |
| Rate limit | Polite — use REQUEST_DELAY = 0.5s between requests |
| Source name | livepocket |
| Source ID format | livepocket_{slug} — URL slug from /e/{slug} path (e.g. livepocket_hmvyt) |
Field Mappings
| Event field | Source element |
|---|
raw_title | h1.heading01 — top-level event title |
start_date | dl.event-detail-info__list → div.event-detail-info__block where dt contains 開催日 → dd text, first YYYY年M月D日 match |
end_date | Same dd text — second date if range 〜 separator present; otherwise same as start_date |
location_name | 会場 dd — text before （都道府県） suffix |
location_address | 会場 dd — <span> after （都道府県）; stripped of map link boilerplate |
raw_description | Composed: 開催日時: header + 会場: + 出演者: (if present) + div.event-detail__content main text |
price_info | div.event-detail-ticket-body — ¥ price strings joined with / |
is_paid | True unless price text contains 無料 or ¥0 / 0円 |
source_id | livepocket_{slug} |
Taiwan Relevance Filter
Applied on full detail-page text (after fetching /e/{slug}).
TAIWAN_KEYWORDS = ["台湾", "Taiwan", "臺灣"]
- Search keywords:
台湾, Taiwan, 臺灣 (三組)
- Filter is applied on the detail page text because search results can match keyword in performer names or unrelated context
- Events with "台湾" only in ticket seller name (e.g. 台湾スタジオ) are still included — broad filter is intentional
CSS Selectors — Critical Note
The dl class is event-detail-info__list, NOT event-detail-info.
dl = soup.select_one("dl.event-detail-info__list")
The dt/dd pairs are wrapped in div.event-detail-info__block divs:
<dl class="event-detail-info__list">
<div class="event-detail-info__block">
<dt class="event-detail-info__title">
<span class="event-detail-info__date">開催日</span>
</dt>
<dd class="event-detail-info__data">2026年7月27日(月)</dd>
</div>
...
</dl>
Use _get_dd_text(dl, label) which iterates div.event-detail-info__block and checks dt.get_text().
There are two identical dl.event-detail-info__list blocks per page (desktop + mobile rendering). Always use select_one() to get the first.
Venue Parsing
Raw venue dd text format:
新宿MARZ (東京都)
東京都新宿区歌舞伎町2-45-1 第一常磐ビルB1
会場マップ・アクセス方法はこちら
location_name: text up to and including (都道府県) parenthetical — matched via _PREF_RE = re.compile(r"[（(][^）)]+[都道府県][）)]")location_address: <span> text after the parenthetical; strip map link boilerplate
Date Extraction Notes
- Single day:
"2026年7月27日(月)" → start_date = end_date
- Date range:
"2026年6月27日(土)〜2026年6月28日(日)" → extract first and second YYYY年M月D日 matches
- Day-of-week
(曜) parentheticals are stripped by the regex matching only \d{4}年\d{1,2}月\d{1,2}日
Search Pagination
Search pages return 404 when the page number exceeds the result count. The scraper breaks the loop cleanly on 404 via _fetch() warning + None return.
Troubleshooting
| Symptom | Likely Cause | Fix |
|---|
| 0 events found | dl.event-detail-info__list selector changed | Inspect page for new dl class; update selector and _get_dd_text block iterator |
start_date null | 開催日 label not found in dt text | Check HTML; the span child text may have changed (e.g. 開催日時) |
location_address null | Venue has no <span> for address (online event or no address listed) | Expected for online events |
| No events pass Taiwan filter | Search results page format changed | Inspect a.event-card[href^='/e/'] selector and badge logic |
| Price info blank | div.event-detail-ticket-body missing | Free events or old ticket format; is_paid falls back to None |
Pending Rules
