Module scrub 

Removing personal data from what the credit registration pipeline keeps of its Suotar exchanges: the call log’s bodies, the ledger’s event details and error messages.

Enums

KeyPolicy 🔒

Constants

NEVER_SCANNED_KEYS 🔒

Keys whose values the value scan must leave alone: they carry ids shaped like student numbers — request item ids, Sisu ids such as hy-CUR-135176012 — that the scan would mangle.

REDACTED

Replaces a redacted value; the key is kept so the payload shape survives.

REDACTED_KEYS 🔒

Keys whose values identify a person or authenticate a request. Matched case-insensitively at any depth.

Statics

EMAIL_RE 🔒

STUDENT_NUMBER_RE 🔒

Student-number-shaped digit runs. The id shapes come first because the regex crate has no lookaround: leftmost-first alternation is the only way to say “digits not part of an id”. The word boundaries keep the digit branch off longer numbers such as millisecond timestamps.

Functions

keep_scalars 🔒

Keeps bare ids and lists of ids, but hands objects back to scrub_suotar_body so an exemption cannot smuggle a nested error message past the value scan.

key_policy 🔒

normalize_key 🔒

scrub_suotar_body

Best-effort removal of personal data from a Suotar request or response body. Not a guarantee and not an exhaustive PII filter.

scrub_text

Scrubs a bare error message with the same rules as a JSON body’s free text.

suotar_exchange_details

Both sides of a Suotar exchange, scrubbed on construction so there is no way to build an unscrubbed details. The request is kept because the ledger row no longer reflects it after a retry.