; Adjust Eshell regression tests to avoid SIGPIPE

In batch mode, SIGPIPEs can cause Emacs to abort (bug#66186).

* lisp/eshell/esh-io.el (eshell-output-object-to-target): Update
comment.

* test/lisp/eshell/esh-proc-tests.el
(esh-proc-test/pipeline-connection-type/middle)
(esh-proc-test/pipeline-connection-type/last): Use '(ignore)', since
that causes no output when called with no arguments, thus avoiding a
risky 'process-send-string'.

1 files changed, 5 insertions, 2 deletions

diff --git a/lisp/eshell/esh-io.el b/lisp/eshell/esh-io.el
index cd0cee6e21d..d0f1e04e925 100644
--- a/lisp/eshell/esh-io.el
+++ b/lisp/eshell/esh-io.el
@@ -648,8 +648,11 @@ Returns what was actually sent, or nil if nothing was sent.")
       (process-send-string target object)
     (error
      ;; If `process-send-string' raises an error and the process has
-     ;; finished, treat it as a broken pipe.  Otherwise, just
-     ;; re-throw the signal.
+     ;; finished, treat it as a broken pipe.  Otherwise, just re-raise
+     ;; the signal.  NOTE: When running Emacs in batch mode
+     ;; (e.g. during regression tests), Emacs can abort due to SIGPIPE
+     ;; here.  Maybe `process-send-string' should handle SIGPIPE even
+     ;; in batch mode (bug#66186).
      (if (process-live-p target)
          (signal (car err) (cdr err))
        (signal 'eshell-pipe-broken (list target)))))