Fix cubic review issues: process_type detection, cmd storage, PID cleanup, and migration

- Fix Process.current() to store psutil cmdline instead of sys.argv for accurate validation
- Fix worker process_type detection: explicitly set to WORKER after registration
- Fix ArchiveResultWorker.start() to use Process.TypeChoices.WORKER consistently
- Fix migration to be explicitly irreversible (SQLite doesn't support DROP COLUMN)
- Fix get_running_workers() to return process_id instead of incorrectly named worker_id
- Fix safe_kill_process() to wait for termination and escalate to SIGKILL if needed
- Fix migration to include all indexes in state_operations (parent_id, process_type)
- Fix documentation to use Machine.current() scoping and StatusChoices constants

Co-authored-by: Nick Sweeting <pirate@users.noreply.github.com>

archivebox/crawls/models.py

@@ -424,9 +424,10 @@ class Crawl(ModelWithOutputDir, ModelWithConfig, ModelWithHealthStats, ModelWith
         if self.OUTPUT_DIR.exists():
             for pid_file in self.OUTPUT_DIR.glob('**/*.pid'):
                 cmd_file = pid_file.parent / 'cmd.sh'
-                # Only delete PID file if kill succeeded or process is already dead
+                # safe_kill_process now waits for termination and escalates to SIGKILL
+                # Returns True only if process is confirmed dead
                 killed = safe_kill_process(pid_file, cmd_file)
-                if killed or not pid_file.exists():
+                if killed:
                     pid_file.unlink(missing_ok=True)
 
         # Run on_CrawlEnd hooks