| name | moonromantic |
| description | Platform rules and field mappings for the 月見ル君想フ (MoonRomantic) live venue scraper |
moonromantic — Source Skill
Platform Overview
- Name: 月見ル君想フ (Tsukimi Ru Kimi Omou / MoonRomantic)
- URL: https://www.moonromantic.com/
- Type: Live music venue (Minami-Aoyama, Tokyo)
- Location: 東京都港区南青山4-9-1 B1F
- Rendering: Wix JS rendering — Playwright required
- Events/month: ~1–2 Taiwan-related events
Scraper Key
--source moon_romantic
Class: MoonRomanticScraper → key auto-derived as moon_romantic (NOT moonromantic).
Strategy
- Build monthly schedule URLs:
https://www.moonromantic.com/allevents/categories/YYYY-MM
- Scrape current month +
MONTHS_AHEAD = 3 future months (4 pages total).
- Collect
/post/{slug} links from each schedule page.
- For each event post, fetch the detail page with Playwright.
- Filter: check Taiwan keywords against
title + body_text.
- Extract date from title prefix
"YYYY.MM.DD | Event Title".
Field Mappings
| Field | Source |
|---|
source_id | moonromantic_{slug_clean} (URL slug, max 80 chars, alphanumeric + hyphens + underscores) |
source_url | Full post URL, e.g. https://www.moonromantic.com/post/260510 |
name_ja | Post title from <h1> (includes date prefix) |
start_date | Parsed from title prefix `"YYYY.MM.DD |
location_name | Always "月見ル君想フ" (hardcoded) |
location_address | Always "東京都港区南青山4-9-1 B1F" (hardcoded) |
raw_description | "開催日時: YYYY年MM月DD日\n\n" + body_text |
Date Parsing
Title format: "2026.05.10 (日) DSPS Japan Tour"
_TITLE_DATE_RE = re.compile(r"^(\d{4})\.(\d{2})\.(\d{2})")
Parses YYYY.MM.DD from start of title.
Taiwan Filter
Most events at this venue are Japanese indie — Taiwan filter is critical.
TAIWAN_KEYWORDS
[
"台湾", "Taiwan", "臺灣", "台灣",
"台北", "高雄", "台中", "台南",
"台日", "日台",
"BIG ROMANTIC", "大浪漫",
"DSPS", "VOOID", "Andr", "Sunset Rollercoaster",
"日落飛車", "告五人", "魚丁糸", "ØZI", "Elephant Gym", "大象體操",
]
Known Issues / Edge Cases
- Slow: Playwright loads 4 monthly pages + N individual post pages. Expect 2–5 min for dry-run.
- Wix lazy-loading: Schedule pages may need
page.wait_for_selector("a[href*='/post/']").
- Post URL skips:
/post/announcement and /post/_system are filtered out.
- Slug max length: 80 chars (truncated if longer) to avoid DB key overflow.
Blocked Post Patterns (Non-Event Reject List)
The following title patterns indicate venue-management / admin posts — they are never Taiwan-related events. Any post whose title matches must be rejected before the Taiwan keyword check:
| Pattern | Reason |
|---|
RENTAL / PRIVATE RENTAL | Venue hire announcement — no cultural content |
RENT (word boundary) | Variant spelling |
場所貸し / 会場貸し | Japanese venue rental terms |
_BLOCKED_POST_PATTERNS = re.compile(
r"\bRENTAL\b|PRIVATE\s+RENTAL|\bRENT\b|場所貸し|会場貸し",
re.IGNORECASE,
)
This block is applied twice:
- Against
first_line of page_text (fast-reject before Taiwan keyword check).
- Against the confirmed
title after extraction (catches nav-renders-first edge cases).
Taiwan Filter — False Positive Warning
The 関連記事 (related articles) sidebar on each post lists other venue events, some of which may be Taiwan-related. Always truncate page_text at 関連記事 before keyword matching. Do NOT use the old > 200 threshold — it was incorrect and caused RENTAL posts to pass when the related section appeared early in the text.
