trim trailing null from callout script get-attributes output

[collinwalling]

A trailing null will make serde_json::from_str fail. Let's trim the string if a null is present to make it happy.

Fixes #84

Note, this will create a merge conflict in PR #83. If this change is acceptable, I can either rebase Boris' patches on top of this, or I can fit this patch in nicely after the conflicting commit.

[jonner]

Consider adding a test while we're at it. For example, something like this could be squashed into your patch:

diff --git a/src/tests.rs b/src/tests.rs
index 215ff2a..a98b5c5 100644
--- a/src/tests.rs
+++ b/src/tests.rs
@@ -1959,6 +1959,16 @@ fn test_callouts() {
             test.populate_callout_script("good-json.sh");
         },
     );
+    test_get_callout(
+        "test_callout_good_json_null",
+        Expect::Pass,
+        Uuid::parse_str(DEFAULT_UUID).unwrap(),
+        DEFAULT_PARENT,
+        DEFAULT_TYPE,
+        |test| {
+            test.populate_callout_script("good-json-null-terminated.sh");
+        },
+    );
     test_get_callout(
         "test_callout_bad_json",
         Expect::Fail(None),
diff --git a/tests/callouts/good-json-null-terminated.sh b/tests/callouts/good-json-null-terminated.sh
new file mode 100755
index 0000000..c5e12d7
--- /dev/null
+++ b/tests/callouts/good-json-null-terminated.sh
@@ -0,0 +1,2 @@
+#!/bin/sh
+echo -ne "[{\"attribute0\": \"VALUE\"}]\0"

[collinwalling]

Fair point. Thanks for providing this.

[collinwalling]

Would you like for me to add your sign-off to the patch since you provided the test case?

[jonner]

I suppose that would probably be more proper, why not.

[jonner]

By the way, I notice that your commit message says "Addresses #84". If you change that to one of the phrases mentioned here (basically: 'closes', 'fixes', or 'resolves'), github will autoclose the issue when we merge this.

[collinwalling]

Fixed up. Thanks @jonner

[jonner]

Let's just merge this and rebase #83 if necessary. That one is going to take more time to review.

[collinwalling]

Sure. Boris is out-of-office for a while. Any suggestions on how I can handle rebasing his branch? Is it as simple as forking his changes, rebasing, then creating a new PR? /shrug.

[jonner]

I think I might have the ability (as maintainer) to edit his PR branch. So maybe I could rebase it and force push. Otherwise we can figure it out later ;)