Replace config-driven HtmlScraperPlugin with specific archive classes

Each archive scraper now has its own class with hardcoded URL and parsing
logic; config only carries auto_queue, timeout, and rate_limit_seconds.

- html_scraper: refactor to base class with public shared utilities
  (YEAR_RE, AUTHOR_PREFIX_PAT, cls_inner_texts, img_alts)
- rusneb.py (new): RusnebPlugin extracts year per list item rather than
  globally, eliminating wrong page-level dates
- alib.py (new): AlibPlugin extracts year from within each <p><b> entry
  rather than globally, fixing nonsensical year values
- shpl.py (new): ShplPlugin retains the dead ШПИЛ endpoint with hardcoded
  params; config type updated from html_scraper to shpl
- config: remove config: subsections from rusneb, alib_web, shpl entries;
  update type fields to rusneb, alib_web, shpl respectively
- plugins/__init__.py: register new specific types, remove html_scraper
- tests: use specific plugin classes; assert all CandidateRecord fields
  (source, title, author, year, isbn, publisher) with appropriate constraints

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

config/functions.default.yaml

@@ -57,28 +57,17 @@ functions:
 
     rusneb:
       name: "НЭБ"
-      type: html_scraper
+      type: rusneb
       auto_queue: true
       rate_limit_seconds: 5
       timeout: 8
-      config:
-        url: "https://rusneb.ru/search/"
-        search_param: q
-        img_alt: true
-        author_class: "search-list__item_subtext"
 
     alib_web:
       name: "Alib (web)"
-      type: html_scraper
+      type: alib_web
       auto_queue: false
       rate_limit_seconds: 5
       timeout: 8
-      config:
-        url: "https://www.alib.ru/find3.php4"
-        search_param: tfind
-        extra_params: {f: "5", s: "0"}
-        encoding: "cp1251"
-        bold_text: true
 
     nlr:
       name: "НЛР"
@@ -91,13 +80,9 @@ functions:
         query_prefix: "title="
 
     shpl:
+      # Endpoint currently returns HTTP 404; retained for future re-enablement.
       name: "ШПИЛ"
-      type: html_scraper
+      type: shpl
       auto_queue: false
       rate_limit_seconds: 5
       timeout: 8
-      config:
-        url: "https://www.shpl.ru/cgi-bin/irbis64/cgiirbis_64.exe"
-        search_param: S21ALL
-        extra_params: {C21COM: S, I21DBN: BIBL, P21DBN: BIBL, S21FMT: briefWebRus, Z21ID: ""}
-        brief_class: "brief"