feat(eat): reset SGR at newline to contain unterminated-color bleed

Claude Code and similar inline TUIs sometimes truncate a colored span without emitting a reset; the unterminated color then bleeds onto every following line in the EAT buffer. Advise eat-term-process-output to inject an SGR reset before each newline, containing the bleed to its own line. Validated in real EAT and via an ansi-color proxy: it contains the bleed and leaves per-line coloring intact, since programs re-open their color on each line. Gated by cj/eat-reset-sgr-at-newline (default on) so it can be disabled if a program ever carries one color across newlines without re-opening it.

1 files changed, 11 insertions, 0 deletions

diff --git a/tests/test-term-tmux-history.el b/tests/test-term-tmux-history.el
index 13e3e1715..a5f5c93cd 100644
--- a/tests/test-term-tmux-history.el
+++ b/tests/test-term-tmux-history.el
@@ -305,5 +305,16 @@ the EAT window line-scrolls instead of recentering on full-frame redraws."
     (should (= scroll-margin 0))
     (should (null auto-window-vscroll))))
 
+(ert-deftest test-term-eat-reset-sgr-at-newline ()
+  "Normal: the SGR-reset advice injects a reset before each newline when enabled
+\(containing an unterminated color), and passes output through unchanged when
+disabled."
+  (let ((cj/eat-reset-sgr-at-newline t))
+    (should (equal (cj/--eat-reset-sgr-at-newline (list (quote term) "a\nb\n"))
+                   (list (quote term) "a\e[0m\nb\e[0m\n"))))
+  (let ((cj/eat-reset-sgr-at-newline nil))
+    (should (equal (cj/--eat-reset-sgr-at-newline (list (quote term) "a\nb\n"))
+                   (list (quote term) "a\nb\n")))))
+
 (provide 'test-term-tmux-history)
 ;;; test-term-tmux-history.el ends here