fix: streamed reply no longer repeats a word when edits fall back to a new message

_continuation_text backed the cut up to the previous space even when the visible
preview already ended on a whole word, so the continuation re-sent that word at
the seam ("the answer" + "answer to your question."). Back up only when the cut
lands inside a word (#116312 behaviour kept).

Found by the platforms E2E suite (#121409).

gateway/stream_consumer_fallback.py

@@ -58,7 +58,9 @@ class StreamFallbackMixin:
             # the continuation re-sends the broken word's tail and reads as an
             # ordinary continuation.  A prefix with no boundary (one very long
             # token) keeps the original cut rather than re-sending the whole reply.
-            if cut < len(final_text):
+            # A prefix that already ends on a whole word needs no back-up: doing it
+            # re-sent that word at the seam.
+            if cut < len(final_text) and not final_text[cut].isspace() and not final_text[cut - 1].isspace():
                 boundary = max(
                     final_text.rfind(" ", 0, cut),
                     final_text.rfind("\n", 0, cut),