edfcore 0.4.472 → 0.4.474

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 = "0.4.472";
+export declare const VERSION = "0.4.474";
 //# 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 = '0.4.472';
+export const VERSION = '0.4.474';
 //# 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.
 
+## 0.4.474
+
+- **Added** all twelve month names to the `dd-MMM-yyyy` subfield tests. `MONTH_ABBREVIATIONS
+  .indexOf(...) + 1` turns a zero-based position into a month number, and `0` doubles as "not a
+  month name" — so January sits exactly on the boundary the bounds check uses.
+- Narrowing `month < 1` to `month <= 1` refuses every January and passed the whole suite. The
+  fixtures ran on August, May and December, and no test had ever read a `JAN` subfield: that is a
+  twelfth of every patient birthdate and every recording `Startdate` there is, on the two fields
+  EDF+ added specifically so a year could be unambiguous.
+- December was covered by accident, through one birthdate fixture. Both ends are now stated rather
+  than left to chance, and the ten in between cost one line.
+- Two non-vacuity cases go with the table: a name that is not one of the twelve is refused, which
+  is what the lower bound is really for, and the numeric spelling is checked at 0, 1, 12 and 13,
+  since it reaches the same bounds by the other route.
+
+## 0.4.473
+
+- **Added** tests for a runtime with no `TextDecoder`, which `tal/grammar.ts` is written for and
+  nothing had ever run. `utf8Decoder()` is two branches — build one, or record that this runtime
+  has none — and every test in the suite runs on Node, where the global is always there. Both the
+  `null` and the message it selects were dead.
+- The degradation is the point of the branch. Annotation text is the one place edfcore decodes
+  UTF-8, and without the fallback a runtime lacking the global would throw while parsing a file
+  that is perfectly valid, losing the onsets, the timeline and every event over text nobody may be
+  reading.
+- What is pinned is the MESSAGE, not just the code. `ANNOTATION_TEXT_NOT_UTF8` also fires for a
+  text run that really is malformed, and the two call for opposite responses: one is a broken
+  file, the other is a working file on a runtime that cannot check it. Both are asserted, each
+  against the other's wording.
+- The Latin-1 expectation is derived from the fixture's own bytes rather than written out, so it
+  cannot drift from the string it is about, and the global is stubbed and restored rather than
+  deleted, so the decoder the module caches cannot leak into another file.
+
 ## 0.4.472
 
 - **Documented and tested** what `threeFields` actually tolerates in a date or time field. Its

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "edfcore",
-  "version": "0.4.472",
+  "version": "0.4.474",
   "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 = '0.4.472';
+export const VERSION = '0.4.474';