[processor/redaction] Add support for allowed values that are never redacted

[janslow]

Component(s)

processor/redaction

Is your feature request related to a problem? Please describe.

I want the processor to never redact specific values, such as email addresses from specific domains. For example, we need [email protected] to be redacted for compliance purposes, but not [email protected]

Describe the solution you'd like

Add an allowed_values configuration option that is a list of regexes (same as blocked_values), then never redact values that match at least one allowed_values regex (even if it also matches a blocked_values regex).

For example, with the given config:

processors:
  redaction:
    # ...
    blocked_values:
      - '.+@.+'
    allowed_values:
      - '[email protected]'

then [email protected] would be redacted in the relevant attributes, but [email protected] wouldn't.

Describe alternatives you've considered

The blocked_values uses Go's regexp package, which doesn't support negative lookahead, which could have been another alternative. There may be workarounds by incrementally excluding matches, but they result in very complex and hard to maintain matches

Additional context

I'm happy to contribute a PR if my proposed solution is acceptable

[github-actions]

Pinging code owners:

• processor/redaction: @dmitryax @mx-psi @TylerHelmuth

See Adding Labels via Comments if you do not have permissions to add labels yourself.