edfcore 0.4.485 → 0.4.487

package/dist/constants.d.ts

@@ -109,5 +109,5 @@ export declare const SIGNAL_FIELD_BLOCK_OFFSETS: {
     readonly reserved: 224;
 };
 /** Published package version. Kept in sync with package.json by a test. */
-export declare const VERSION = "0.4.485";
+export declare const VERSION = "0.4.487";
 //# sourceMappingURL=constants.d.ts.map

package/dist/constants.js

@@ -79,5 +79,5 @@ export const SIGNAL_FIELD_BLOCK_OFFSETS = {
     reserved: 224,
 };
 /** Published package version. Kept in sync with package.json by a test. */
-export const VERSION = '0.4.485';
+export const VERSION = '0.4.487';
 //# sourceMappingURL=constants.js.map

package/docs/CHANGELOG.md

@@ -6,6 +6,39 @@ alone does not tell you whether you were affected.
 edfcore is pre-1.0. Patch releases have carried behaviour changes where the old behaviour was a
 defect; those are called out below.
 
+## 0.4.487
+
+- **Set** the suite's test timeout explicitly, at thirty seconds, with the reasoning in the config.
+  vitest's default is five, which nobody here chose, and no test in this suite uses the clock as an
+  assertion — the cost tests count READS and BYTES, which is what makes them stable. The only job
+  a timeout has here is stopping an infinite loop from taking the run with it.
+- Five seconds is too tight for that job. Several tests build multi-megabyte fixtures and the
+  heaviest ordinary case runs in about a second, so the default left under six times its own cost
+  as headroom — and a run under `--coverage` is several times slower, which is where the failures
+  appeared: two in `read-pattern.test.ts` and one in `envelope.test.ts`, all three passing on the
+  same commit without instrumentation. A full coverage run of the suite now passes where it failed
+  before this change.
+- A timeout that fires on correct work teaches the reader to rerun rather than to read, which is
+  the opposite of what a red test is for.
+- Thirty seconds is thirty times the slowest ordinary test and still fails a genuine hang well
+  inside a minute. The files needing longer still say so themselves, and 0.4.484's `beforeAll` is
+  unaffected: moving a file-level cost out of an arbitrary test body is right whatever the budget
+  is.
+
+## 0.4.486
+
+- **Added** tests for the sentence the repeat counter turns into. `parseTalRegion` collapses many
+  occurrences of one defect into a single issue carrying a count, and `grammar.test.ts` pins the
+  counter thoroughly — but nothing checked whether the clause `reportIssue` builds from it appears
+  when it should and stays away when it should not.
+- Relaxing that guard puts "(1 occurrences in this region; the first is described)" on the
+  ordinary case. Ungrammatical, and worse than ungrammatical: it tells a reader looking at a
+  single bad TAL that there are others they cannot see. The same defect 0.4.421 fixed on the
+  validation report's first line — a count rendered without asking whether it needed rendering.
+- Both directions and the threshold: one occurrence carries no count, two says two, three says
+  three, and the bytes quoted are the FIRST offender's, which is what the clause promises. The
+  two-occurrence case is what stops a guard reading `> 2` passing everything else.
+
 ## 0.4.485
 
 - **Added** the edge cases for the sub-second start offset, which is documented as `[0, 1)` and

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "edfcore",
-  "version": "0.4.485",
+  "version": "0.4.487",
   "description": "Modern, typed, zero-dependency reader for EDF, EDF+, BDF and BDF+ biosignal files. Works in browsers and Node with true random access.",
   "keywords": [
     "edf",

package/src/constants.ts

@@ -93,4 +93,4 @@ export const SIGNAL_FIELD_BLOCK_OFFSETS = {
 } as const;
 
 /** Published package version. Kept in sync with package.json by a test. */
-export const VERSION = '0.4.485';
+export const VERSION = '0.4.487';