use the tools from pipeline_data.rs to print true tables

This commit uses the construct from `PipelineData::print` to
- get the "declaration ID" of the `table` command
- execute that command on the example outputs

Then we use the construct from `PipelineData::write_all_and_flush`
to
- iterate over all the items of the table
- print them in a pretty way

.# Example output
```nushell
> help merge
...
  Add an 'index' column to the input table
  > [a b c] | wrap name | merge ( [1 2 3] | wrap index )
  ╭───┬──────╮
│ # │ name │
├───┼──────┤
│ 1 │ a    │
│ 2 │ b    │
│ 3 │ c    │
╰───┴──────╯
```

.## Limitations
- the format is not perfect for now
- the use of private method bodies from `PipelineData` is not ideal

crates/nu-engine/src/documentation.rs

@@ -213,11 +213,22 @@ fn get_documentation(
 
         match &example.result {
             Some(result) => {
-                for item in PipelineData::Value(result.clone(), None) {
+                let decl_id = engine_state.find_decl("table".as_bytes(), &[]).unwrap();
+                let table = engine_state
+                    .get_decl(decl_id)
+                    .run(
+                        engine_state,
+                        stack,
+                        &Call::new(Span::new(0, 0)),
+                        PipelineData::Value(result.clone(), None),
+                    )
+                    .unwrap();
+
+                for item in table {
                     let _ = write!(
                         long_desc,
-                        "{}\n",
-                        item.into_string("\n", engine_state.get_config())
+                        "  {}\n",
+                        item.into_string("\n  ", engine_state.get_config())
                     );
                 }
             }