Unable to access JSON keys containing spaces using add_entries - value_expression, rename_keys, or copy_values processors

We are encountering an issue in Data Prepper when trying to access or manipulate JSON fields whose key names contain spaces.

Example Input

{
      "Computer Name": "xyz"
}

Attempt 1: add_entries with value_expression

- add_entries:
    entries:
      - key: cloud/provider
        value_expression: "/Computer Name"

Observed behavior:

Using add_entries with value_expression results in a runtime error:

Error adding entry to record ... value_expression [/Computer Name]

When using add_entries with format instead of value_expression, no error is thrown, but the resulting value is null.

Attempt 2: rename_keys and copy_values

Both processors fail when referencing the same key with a space.

Observed errors:

java.lang.IllegalArgumentException:
key Computer Name must contain only alphanumeric chars with .-_@/
and must follow JsonPointer (ie. 'field/to/key')

And during pipeline validation:

Parameter "entries.null.from_key" for plugin "rename_keys" is invalid:
key Computer Name must contain only alphanumeric chars with .-_@/
and must follow JsonPointer (ie. 'field/to/key')

Expected Behavior

Data Prepper should either:

Support accessing JSON keys containing spaces using valid JSON Pointer syntax, or

Clearly document the limitation and provide a supported workaround for handling such keys (e.g., escaping, bracket notation, or preprocessing support).

Impact

This prevents ingestion and normalization of source data that contains valid JSON keys with spaces, which is common in several log and event sources.

Start by reproducing the reported configurations for add_entries, rename_keys, and copy_values with the example JSON key "Computer Name". Compare the runtime and validation failures with the processors' supported key syntax, then establish whether valid JSON Pointer access or a documented workaround is the intended resolution.