From: Barret Rhoden <brho@google.com>
Date: Tue, 04 Jun 2019 20:41:33 GMT
Subject: Re: Git Test Coverage Report (Thursday, May 30th)
Message-ID: <09bc8c7a-bab5-7cdb-75dc-f7c45a4dbaf8@google.com>
In-Reply-To: <9ab619bb-9deb-4e57-a3ad-9e996425b783@google.com>

Hi -

On 6/4/19 12:38 PM, Barret Rhoden wrote:
> I'll try to come up with a test, independent of my blame-ignore work, 
> that can get in that block.

I have a test that covers blame_coalesce(), which works both with and 
without my blame-ignore commit that started this thread.

However, the only thing we are really testing is that git blame didn't 
crash.  There is no detectable change to the output.  AFAIK, 
blame_coalesce() is a performance enhancement.

If you all are interested in that sort of test, I can put it in a patch. 
  Right now, I have this (below).

Thanks,

Barret


diff --git a/t/t8003-blame-corner-cases.sh b/t/t8003-blame-corner-cases.sh
index c92a47b6d5b1..4c652b85a55b 100755
--- a/t/t8003-blame-corner-cases.sh
+++ b/t/t8003-blame-corner-cases.sh
@@ -275,4 +275,32 @@ test_expect_success 'blame file with CRLF 
core.autocrlf=true' '
  	grep "A U Thor" actual
  '

+test_expect_success 'blame coalesce' '
+	cat >giraffe <<-\EOF &&
+	ABC
+	DEF
+	EOF
+	git add giraffe &&
+	git commit -m "original file" &&
+	
+	cat >giraffe <<-\EOF &&
+	ABC
+	XXX
+	DEF
+	EOF
+	git add giraffe &&
+	git commit -m "interior XXX line" &&
+	
+	cat >giraffe <<-\EOF &&
+	ABC
+	DEF
+	EOF
+	git add giraffe &&
+	git commit -m "same contents as original" &&
+
+	git blame giraffe &&
+
+	true
+'
+
  test_done