[Lldb-commits] [lldb] Add new Python API `SBCommandInterpreter::GetTranscript()` (PR #90703)

[via lldb-commits]

================
@@ -85,3 +86,84 @@ def test_command_output(self):
         self.assertEqual(res.GetOutput(), "")
         self.assertIsNotNone(res.GetError())
         self.assertEqual(res.GetError(), "")
+
+    def test_structured_transcript(self):
+        """Test structured transcript generation and retrieval."""
+        # Get command interpreter and create a target
+        self.build()
+        exe = self.getBuildArtifact("a.out")
+
+        target = self.dbg.CreateTarget(exe)
+        self.assertTrue(target, VALID_TARGET)
+
+        ci = self.dbg.GetCommandInterpreter()
+        self.assertTrue(ci, VALID_COMMAND_INTERPRETER)
+
+        # Send a few commands through the command interpreter
+        res = lldb.SBCommandReturnObject()
+        ci.HandleCommand("version", res)
+        ci.HandleCommand("an-unknown-command", res)
+        ci.HandleCommand("breakpoint set -f main.c -l %d" % self.line, res)
+        ci.HandleCommand("r", res)
+        ci.HandleCommand("p a", res)
+        total_number_of_commands = 5
+
+        # Retrieve the transcript and convert it into a Python object
+        transcript = ci.GetTranscript()
+        self.assertTrue(transcript.IsValid())
+
+        stream = lldb.SBStream()
+        self.assertTrue(stream)
+
+        error = transcript.GetAsJSON(stream)
+        self.assertSuccess(error)
+
+        transcript = json.loads(stream.GetData())
+
+        # The transcript will contain a bunch of commands that are run
+        # automatically. We only want to validate for the ones that are
+        # listed above, hence trimming to the last parts.
+        transcript = transcript[-total_number_of_commands:]
----------------
royitaqi wrote:

*There is a lot more*. I am not sure entirely sure if it's from the set up code in the above. It seems they contain alias or custom command setup. I don't have a completely list. If you want I can print you one here in comment.

The point is that the unit test shouldn't care about any extra command that was ran that is irrelevant to this test. So trim makes sense to me.

https://github.com/llvm/llvm-project/pull/90703