Replace wget/SingleFile/Playwright backends with Zimit

- Zimit Docker container handles all site types (static, SPA, JS redirects)
- Removed: _detect_crawl_mode, _crawl_wget, _crawl_singlefile, preflight logic
- Added: _crawl_zimit() with Docker lifecycle management
- Simplified pipeline: submit → Zimit crawl → kiwix-manage register → done
- No more zimwriterfs step — Zimit produces ZIM directly
- Dashboard UI simplified: removed crawl mode dropdown
- Config simplified: removed reject patterns, preflight, singlefile sections

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

config.yaml

@@ -414,81 +414,12 @@ peertube:
   poll_interval: 1800                    # Seconds between PeerTube acquisition polls (30 min)
 
 scraper:
-  workspace: /opt/recon/data/scraper      # Working directory for wget mirrors + ZIM builds
+  workspace: /opt/recon/data/scraper      # Working directory (tmp dirs for Zimit output)
   output_dir: /mnt/kiwix                  # Finished .zim files land here (kiwix-serve library)
-  rate_limit_delay: 0.5                   # Seconds between wget requests (--wait)
-  wait_random: 1.0                        # Random jitter added to wait (--random-wait range)
   default_language: eng                   # ISO 639-3 language code for ZIM metadata
-  user_agent: "Mozilla/5.0 (compatible; RECON/1.0; +https://echo6.co)"
   poll_interval: 300                      # Seconds between checking for pending scrape jobs
-  keep_workspace_on_failure: true         # Retain workspace for debugging when a job fails
-
-  # Default URL patterns rejected by wget --reject-regex.
-  # Covers common CMS junk across WordPress, Squarespace, Wix, Ghost, Drupal, etc.
-  # Per-job overrides: additional_reject_patterns (appended) or skip_default_patterns (bypass).
-  default_reject_patterns:
-    # WordPress
-    - '\?share='
-    - '\?replytocom='
-    - '\?like_comment='
-    - '/feed/'
-    - '/wp-json/'
-    - '/wp-login'
-    - '/wp-admin'
-    - '/wp-cron'
-    - '\?attachment_id='
-    - '/xmlrpc'
-    - '/trackback'
-    - '/comment-page-'
-    - '\?doing_wp_cron'
-    # Squarespace
-    - '\?format=json'
-    - '\?format=rss'
-    - '/api/'
-    # Wix
-    - '/_api/'
-    - '/_partials/'
-    # Ghost
-    - '/ghost/'
-    - '/p/'
-    # Drupal
-    - '\?q=comment'
-    - '\?q=node'
-    - '/user/login'
-    - '/user/register'
-    # General CMS / site chrome
-    - '/login'
-    - '/signup'
-    - '/register'
-    - '/cart'
-    - '/checkout'
-    - '/search\?'
-    - '/tag/'
-    - '/author/'
-    - '\?print='
-    - '\?pdf='
-    - '\?format=amp'
-    - '\?preview='
-    - '/rss'
-    - '/atom'
-    - '/cdn-cgi/'
-
-  # Pre-flight mode detection
-  preflight:
-    enabled: true
-    timeout: 30                    # Seconds for single-page Playwright fetch
-    min_static_size: 5120          # 5KB - wget HTML below this = suspect JS site
-    min_browser_size: 20480        # 20KB - browser HTML above this confirms JS
-    spa_markers:
-      - 'div#root'
-      - 'div#app'
-      - 'div#__next'
-
-  # SingleFile CLI settings (browser crawl mode)
-  singlefile:
-    executable: single-file
-    chromium_path: "/usr/bin/chromium-browser"
-    crawl_max_depth: 10
+  docker_image: ghcr.io/openzim/zimit     # Zimit Docker image for web crawling
+  docker_workers: 2                       # Concurrent crawl workers inside Zimit container
 
 # Stream B: New Library Pipeline
 new_pipeline: