fix(nextflow): don't mark workspace scan flushed when both probes fail

_flush_deferred_workspace_scan sets _workspace_scan_flushed unconditionally after its
two completion() flush attempts, even when both raised. A later find_referencing_symbols
call then silently skips the flush and answers from a possibly-incomplete workspace AST
cache for the rest of the session, with no exception and no retry.

Fixes #1871

src/solidlsp/language_servers/nextflow_language_server.py

@@ -320,12 +320,14 @@ class NextflowLanguageServer(SolidLanguageServer):
             "textDocument": {"uri": self._resolve_file_uri(relative_file_path)},
             "position": {"line": 0, "character": 0},
         }
+        flushed = False
         for _ in range(2):
             try:
                 self.server.send.completion(params)
+                flushed = True
             except Exception as e:
                 log.debug("Completion request used to flush the Nextflow workspace scan failed: %s", e)
-        self._workspace_scan_flushed = True
+        self._workspace_scan_flushed = flushed
 
     @override
     def _send_references_request(self, relative_file_path: str, line: int, column: int) -> list[lsp_types.Location] | None: