When “nbdkit ... --run 'nbdinfo ...' > out” is expected to fail, we cannot be sure of the content of the "out" file. In particular nbdinfo produces its output incrementally, so in JSON mode there's usually a stray "{" in the output, and in non-JSON mode it usually prints the "protocol: " line. Thus it's not correct to test for the output file being non-empty. Remove these lines which caused the test to fail. However I replaced them with "cat" for diagnostic purposes. Fixes: commit 7623b2cc45078cca88fdd2d96c70c7f82a0db49d --- tests/test-exportname.sh | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tests/test-exportname.sh b/tests/test-exportname.sh index 2ec45023..54e04f86 100755 --- a/tests/test-exportname.sh +++ b/tests/test-exportname.sh @@ -172,12 +172,12 @@ test "$(jq -c "$query" exportname.out)" = '[["c","cc",3]]' # Test strict mode nbdkit -U - --filter=exportname sh exportname.sh exportname-strict=true \ --run 'nbdinfo --no-content --json "$uri"' > exportname.out && fail=1 -test ! -s exportname.out +cat exportname.out nbdkit -U - --filter=exportname sh exportname.sh exportname-strict=true \ exportname=a exportname=b exportname=c \ --run 'nbdinfo --no-content --json "$uri"' > exportname.out && fail=1 -test ! -s exportname.out +cat exportname.out nbdkit -U - --filter=exportname sh exportname.sh exportname-strict=true \ exportname=a exportname=b exportname= default-export=a\ -- 2.27.0