From: Sylvestre Ledru <sylvestre@debian.org>
Date: Sat, 27 Jun 2026 09:02:51 +0000 (+0200)
Subject: tests: shuf: check -o output is preserved when the run fails
X-Git-Url: http://git.ipfire.org/gitweb.cgi?a=commitdiff_plain;h=970548101e9bed4a3ade5d8236e311a08ba0bf5c;p=thirdparty%2Fcoreutils.git

tests: shuf: check -o output is preserved when the run fails

A failed shuf invocation should leave an existing output file untouched
rather than truncating it before the failure is detected.
* tests/shuf/shuf.sh: Verify that a missing input file or an unreadable
random source leaves an existing -o file unchanged.

Link: https://github.com/coreutils/coreutils/pull/304
---

diff --git a/tests/shuf/shuf.sh b/tests/shuf/shuf.sh
index 387bbb649c..86a44831ea 100755
--- a/tests/shuf/shuf.sh
+++ b/tests/shuf/shuf.sh
@@ -136,6 +136,17 @@ done
 { shuf -i0-9 --random-source A --random-source B || test $? -ne 1; } &&
   { fail=1; echo "shuf did not detect multiple --random-source usage.">&2 ; }
 
+# A failed run must leave an existing -o output file untouched rather than
+# truncating it before the failure is detected.
+printf 'precious\n' > exp || framework_failure_
+cp exp out || framework_failure_
+returns_ 1 shuf -o out missing-input || fail=1
+compare exp out || { fail=1; echo "shuf truncated -o on input error">&2 ; }
+cp exp out || framework_failure_
+returns_ 1 shuf -o out --random-source=missing-input in || fail=1
+compare exp out ||
+ { fail=1; echo "shuf truncated -o on random-source error">&2 ; }
+
 # Test --repeat option
 
 # --repeat without count should return an indefinite number of lines