edfcore 0.4.365 → 0.4.367

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 = "0.4.365";
+export declare const VERSION = "0.4.367";
 //# 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 = '0.4.365';
+export const VERSION = '0.4.367';
 //# sourceMappingURL=constants.js.map

package/docs/CHANGELOG.md

@@ -6,6 +6,40 @@ 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.
 
+## 0.4.367
+
+- **Extended** 0.4.338's interface check to the three structural shims `api-sources.md` prints and
+  to the adapter table under them. The shims exist so neither the DOM nor `@types/node` leaks into
+  the published `.d.ts` — naming `Blob` forces `lib.dom` on every consumer — and each is the
+  minimum shape edfcore uses.
+- `shim-assignability.test-d.ts` already proves a `Blob`, a `File`, an `AbortSignal` and a
+  `Response` all fit. What it cannot see is whether the page prints the same shape it proves things
+  about, and a shim is exactly the declaration someone copies off a page to build a test double
+  from: a member the page shows and the type does not is a double that compiles against the page
+  and is refused by the library.
+- `FetchLike` is checked for an absence instead. `signal` is deliberately not in its printed `init`
+  and is still passed at runtime, because naming it would break the assignability of
+  `globalThis.fetch` — so the check is that it stays absent from both, and that the page still
+  explains why, since without the explanation the omission reads as an oversight.
+- The adapter table's rows now have to name real exports from the entry point each row names, with
+  the two filesystem adapters behind `edfcore/node` and everything else reachable from the
+  universal entry — the split the shims exist to protect.
+
+## 0.4.366
+
+- **Added** a check that every version number this repository cites is one it released. The
+  codebase explains itself in release numbers — docblocks say "fixed in 0.3.56", the documentation
+  pages date every behaviour they describe, the changelog cross-references itself — and there are
+  several hundred such citations across 248 files, none of which had ever been checked.
+- They rot silently in two directions: a transposed digit points a reader at nothing, and a
+  citation written while a version is still being cut names a number that never shipped, which is
+  the same failure the fourteen changelog holes were. A wrong version number looks exactly like a
+  right one to a compiler, a linter and a human skimming a diff.
+- The hard part is telling a version from everything else shaped like one, since this tree is full
+  of EDF startdates and starttimes, spec clauses like `2.1.1` and runtime floors like `22.12.0`. A
+  loose scan reports forty-five of those and no real defects; `0.` with no leading zero separates
+  them exactly, and that narrowing is asserted against the lookalikes rather than left to luck.
+
 ## 0.4.365
 
 - **Added** a check that the layer table in `AGENTS.md` names the modules that are actually at each

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "edfcore",
-  "version": "0.4.365",
+  "version": "0.4.367",
   "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 = '0.4.365';
+export const VERSION = '0.4.367';