Appending cell to the existed notebook

The `nbf.write()` seems working pretty good with new notebook generation.
```
import nbformat as nbf
nb = nbf.v4.new_notebook()

text = """\
# Manual EDA with automatic notebook genration"""

code = """\
%pylab inline
hist(normal(size=2000), bins=50);"""

nb['cells'] = [nbf.v4.new_markdown_cell(text),
nbf.v4.new_code_cell(code)
]

with open('test.ipynb', 'a') as f:
nbf.write(nb, f)
```

But I couldn't find the same smooth method in `nbformat` to append cells to existed notebook.
Is there a simpler way to that?

```
import json
with open('eda.ipynb', 'r') as f:
json_obj = json.load(f)

# concat cells
json_obj['cells'] = json_obj['cells'] + nb['cells']

with open('test.ipynb', 'w') as f:
json.dump(json_obj, f)
```

Start by reviewing the nbformat read and write APIs, especially the nbf.write() entry point and how notebook cells are represented. Determine whether appending cells to an existing .ipynb file can be supported through a public API rather than manual JSON handling. Done means the supported approach preserves the existing notebook and adds the new cells.

Written by the indexing model from the issue text.