Context
T169 and T171 provide finite obstructions for arbitrary hidden source languages but no computable radius. T173 shows that regularity turns the missed-word relation for each candidate family into a finite automaton with an explicit size bound.
Hypotheses and scope
- The hidden source A is recognized by a complete r-state DFA.
- The projection is letter-to-letter between finite alphabets.
- The family has m total deterministic synchronous selectors with at most s states each.
Proof or evidence
A missed-word NFA stores one nondeterministically chosen source run and, for each selector, its control together with a source-DFA run on its emitted lift. It has at most r^(m+1)s^m states and recognizes exactly pi(A) minus the union of the selected acceptance domains. A shortest accepting run is simple and therefore has length below the state bound.
Verification notes
The checker exhausts all 1,024 two-state source DFAs, 10,240 one-state family cases of widths one and two, and 262,144 two-state singleton cases. It checks 4,287,052 product transitions and directly validates 105,532 shortest missed words.
Limitations
- The product-state horizon is uniform but need not be tight.
- The bounded-budget decision procedure is finite rather than efficient.
- The theorem does not cover asynchronous selectors or nonregular hidden sources.
- It gives no effective horizon for the deterministic visibly pushdown lift used in FPRD-T168.