Flatten and Rebuild Nested Records

Task

Define this plain record, then write flatten_records(records, encode_value) and rebuild_records(rows, decode_value):

Each record ID is unique. data is built from dictionaries, lists, and scalar values. Dictionary keys are strings. encode_value(value) converts one scalar to text; decode_value(text) reverses that conversion.

The flat form contains one tuple for every node:

(record_id, path, kind, value)

path is a tuple of steps from the record's data root. A dictionary step is ("key", key) and a list step is ("index", index). kind is "dict", "list", or "value". Container rows use None for value; scalar rows use encode_value. Emit each root first, then visit dictionary values in insertion order and list values from left to right.

rebuild_records receives rows in that order and returns the original record shape. The supplied rows are valid, contain one root per record, and never refer to a missing parent. Preserve record order and do not modify an input.

Example

records = [NestedRecord("r1", {"scores": [4, 7], "note": "ok"})]

The first three flat rows describe the root dictionary, the scores list, and its first value:

Empty dictionaries and lists still receive their own row, so rebuilding does not lose them.

Your implementation

Edit solution.py, keep the record definition above, and keep these function signatures:

Do not import a serialization library, modify an input, print, or ask for input. The helper functions own scalar serialization; your code owns paths, container shape, record identity, and order.