7 Commits

Author	SHA1	Message	Date
lilleman	f0305f191f	Re-pin the seeded outputs now that a one-item choice draws nothing	2026-09-02 12:39:04 +02:00
lilleman	de8c1d13e9	Tests for one spelling per shape: no one-item choice, repeated item or inert object	2026-09-02 12:38:07 +02:00
lilleman	99d8aa0fb7	Tests for literal format text, brace escapes and the class builtins	2026-09-02 12:28:51 +02:00
lilleman	04b7b6a432	Rename the exported type to Generator and retire the faker wording	2026-09-01 21:34:27 +02:00
lilleman	05dc042e49	Rename the project to fejkdata and record the fork source	2026-09-01 20:24:03 +02:00
M	1ba9185edc	Pin the calc stream against a moving operand ... The calc case drew from single-element lists, so its pinned bytes were constant wherever the rng sat — it could not have caught a change to how many draws a calc consumes, which is exactly what the operand hold changed. Multi-element lists make that stream a pinned fact, and every pinned pair multiplies correctly, so the hold is pinned with it. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-31 21:58:37 +02:00
lilleman	5282f37a59	Golden test pinning the seeded output of each part of the token grammar	2026-08-28 13:20:07 +02:00