Fast Regex Fallback

[sidmohan0]

• Story 1.1 – Design fallback spec

  • Enumerate MVP entities (email, phone, SSN, credit-card, IPv4/6, DOB, ZIP).
  • Define return schema (same Pydantic model spaCy uses).

• Story 1.2 – Implement RegexAnnotator ← blocked-by 1.1

  • Write hardened regex patterns.
  • Create RegexAnnotator.annotate(text) wrapper.
  • Add unit tests for each entity and edge cases.

• Story 1.3 – Integrate into TextService ← blocked-by 1.2

  • Add engine="regex" | "spacy" | "auto" flag.
  • Default "auto": regex first, fall back to spaCy when needed.
  • Preserve behaviour when users ask explicitly for spaCy.

• Story 1.4 – Performance guardrail ← blocked-by 1.3

  • Add pytest-benchmark test for 1 kB sample (< 20 µs target).
  • Fail CI if runtime > 110 % of saved baseline.