edfcore 0.4.451 → 0.4.453

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 = "0.4.451";
+export declare const VERSION = "0.4.453";
 //# 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 = '0.4.451';
+export const VERSION = '0.4.453';
 //# 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.
 
+## 0.4.453
+
+- **Added** tests for the date comparison behind `DATE_IMPLAUSIBLE`, at every level rather than
+  the year alone. The half of that code which is reachable reports a patient born after the
+  recording that captured them — not a hypothetical shape, since EDF's start date is two digits and
+  the EDF+ rule that 00..84 mean 2000..2084 turns a 1974 recording into a 2074 one.
+- The comparison is three lines — year, then month, then day — and every fixture that reached it
+  differed in the year, so two of the three had never run. A comparison stopping at the year still
+  catches the case that motivated the check, and silently stops catching a birthdate later in the
+  same year: a mistyped month, or a `dd-MMM-yyyy` subfield with its day and month transposed, which
+  is every date before the thirteenth.
+- The equality boundary is the other half and the one worth getting right in this direction. A
+  recording made on the day of birth is a neonatal EEG, which is an ordinary thing to point this
+  library at; `> 0` rather than `>= 0` is what keeps that file clean, and an off-by-one there would
+  report a defect on every recording taken in a delivery suite.
+
+## 0.4.452
+
+- **Added** tests that a server ignoring Range is downloaded once however the readers arrive.
+  `allowFullDownload: true` turns every read into a claim on one transfer, and getting that wrong
+  is not slow but fatal: every read that entered `fetchRange` used to issue its own GET, each
+  buffering the whole resource, so N concurrent block reads downloaded the file N times and held up
+  to `maxConcurrency` copies at once.
+- `hardening.test.ts` pins the count for readers arriving together at the default concurrency.
+  Three other ways to arrive each take a different branch: queued behind the gate with
+  `maxConcurrency: 1`, where waiting for a slot is exactly when a stale decision is most likely;
+  arriving while the first reader is still finding out whether the server honours Range; and
+  arriving once everything has settled, where nothing should go out at all.
+- Requests and transfers are counted separately, because different guards protect them. The one
+  inside the response handler stops a second download; the ones before the request stop a second
+  GET. With only the first, eight readers cost eight requests and one body — the same bill on a
+  metered origin, for the same bytes, and nothing in the result to show it.
+
 ## 0.4.451
 
 - **Added** a property test for `bucketCount` over arbitrary windows. `api-helpers.md` calls it

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "edfcore",
-  "version": "0.4.451",
+  "version": "0.4.453",
   "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 = '0.4.451';
+export const VERSION = '0.4.453';