waseda-icl

name	waseda_icl
description	Platform rules, WP REST API date extraction, and Taiwan filter conventions for the waseda_icl scraper
applyTo	scraper/sources/waseda_icl.py

Waseda ICL Scraper — Platform Reference

Platform Profile

Field	Value
Site URL	https://www.waseda.jp/folaw/icl/news/
API/Rendering	WordPress REST API (/wp-json/wp/v2/posts?search=台湾)
Auth required	No
Rate limit	None observed
Source name	waseda_icl
Source ID format	waseda_icl_{wp_post_id} (integer, stable)

Key Differences from waseda_taiwan.py

• Different site: waseda.jp/folaw/icl/ (比較法研究所) vs waseda-taiwan.com
• NOT Cloudflare-blocked: ICL subdirectory allows WP REST API; main waseda.jp does not
• Separate post IDs: Events co-hosted with waseda-taiwan.com are posted on both sites with distinct WP post IDs — no source_id collision
• Frequency: ~1–2 events per year (legal symposia, Japanese-Taiwan law lectures)

Field Mappings

Event Field	Source
source_id	waseda_icl_{wp_post_id}
start_date	【開催日時】 / 【日　時】 labels in post content
location_name	【開催会場名】 / 【場　所】 labels in post content
raw_description	"開催日時: YYYY年M月D日\n\n" + stripped post content
name_ja	WP post title

Taiwan Relevance Filter

• Two-stage: WP API search (search=台湾 + search=Taiwan) → title must also match _TAIWAN_TITLE_RE
• Title keywords: 台湾, 日台, Taiwan, 臺灣
• Skip 開催報告: Posts with 【開催報告】 or 開催されました in title are already-past events — always skip

Date Extraction

• Labels inside 【】: 【開催日時】, 【日　時】 (with full-width space)
• _extract_after_bracket_label() stops at the next 【 or end of string
• No year in some older posts: infer from publication date if YYYY not found in content
• LOOKBACK_DAYS = 120 — covers 4 months back

Troubleshooting

Symptom	Likely Cause	Fix
0 events	No Taiwan events in current window (expected ~1–2/year)	Check manually at /news/ — 0 is normal
AttributeError on date	Post has date label but malformed format	Add format to _JAPANESE_DATE_RES
Scraper not running	Not registered in SCRAPERS — see history	Run SCRAPERS audit

waseda_icl-specific Rules

1. _REPORT_RE skip is critical: Without it, all past events (事後報告 posts) would match. Always skip posts where title matches 【開催報告】 or が開催されました.
2. WP API search is not exact match: search=台湾 also returns posts where 台湾 appears only in the body — hence the secondary _TAIWAN_TITLE_RE title-level filter.
3. 0 events is almost always correct: This is a legal research institute; genuine Taiwan events occur ~1–2 times per year. Do not add more keywords to inflate results.
4. LOOKBACK_DAYS = 120: Longer than most scrapers because event announcements may be posted up to 3 months before the event date.
5. No overlap with waseda_taiwan: Different institutions; do not merge them via merger.py.

Pending Rules