[Lldb-commits] [lldb] [lldb-dap] Add clipboard context support (PR #170644)

[John Harrison via lldb-commits]

================
@@ -836,7 +837,10 @@ std::string VariableDescription::GetResult(protocol::EvaluateContext context) {
   // Try the SBValue::GetDescription(), which may call into language runtime
   // specific formatters (see ValueObjectPrinter).
   lldb::SBStream stream;
-  v.GetDescription(stream);
+  if (context == protocol::eEvaluateContextRepl)
----------------
ashgti wrote:

I think that `eDescriptionLevelFull` makes sense for the clipboard context.

For reference, `vscode-js-debug` will JSONify the result when you copy the value, so given `let a = {foo: "bar", baz: { a: [1, 2, 3, 4, 5, 6, 7, 8, 9], b: 123n, "complex key": true }}` `evaluate({'expression': 'a', 'context': 'clipboard'})` becomes:

```
{
  foo: "bar",
  baz: {
    a: [
      1,
      2,
      3,
      4,
      5,
      6,
      7,
      8,
      9,
    ],
    b: 123n,
    "complex key": true,
  },
}
```

I could see the repl context being somewhat less verbose if we have the variable reference for the value, since you can expand the variable reference to dive into details under the top level result.

But, I think either of these are somewhat subjective so feel free to go with your preference.

https://github.com/llvm/llvm-project/pull/170644