edfcore 0.4.302 → 0.4.304

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 = "0.4.302";
+export declare const VERSION = "0.4.304";
 //# 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 = '0.4.302';
+export const VERSION = '0.4.304';
 //# sourceMappingURL=constants.js.map

package/docs/CHANGELOG.md

@@ -6,6 +6,32 @@ 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.
 
+## 0.4.304
+
+- **Executed** the costing on `large-files.md`. Its whole argument is numeric: an eight-hour,
+  30-channel, 256 Hz EDF — 28,800 one-second records of 15,360 bytes — where a ten-second window
+  is one read of 153,600 bytes out of 442,375,936, or 0.035 % of the file, opening it costs 7,936
+  bytes, and asking for one channel out of thirty costs byte-for-byte the same. That is the
+  random-access claim stated as money, and every figure was prose.
+- They are the numbers a reader checks their own instinct against. Someone who expects
+  `signalIndices: [0]` to be thirty times cheaper needs the page to be right about it, because the
+  advice to name every channel in one call rests on that. The header is built at full width and
+  the arithmetic checked against what edfcore reports for it — record size, the byte offset the
+  window lands at, the last byte the read touches, and the overread factor of 30, which is the
+  record over one signal's block.
+
+## 0.4.303
+
+- **Executed** the request budget on `api-sources.md`: one `HEAD` for the length, `bytes=0-255`,
+  one more range for the rest of the header, and one whole record at each end for the timekeeping
+  probes — five in total, with the caller's headers on all five. That is the paragraph a reader
+  consults before pointing this at S3, and every clause is a cost they are budgeting.
+- The count was pinned elsewhere against a literal; the composition was prose, and so was the
+  promise about headers — the clause with a security shape, since one request quietly going out
+  without the configured `Authorization` 403s in production and nowhere else. All of it is now
+  driven through an injected `fetch`, which the suite requires of anything touching the network
+  and which is also how the page tells a reader to test their own adapter.
+
 ## 0.4.302
 
 - **Made** `api-reading.md`'s read counts the expectation rather than a second statement of it.

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "edfcore",
-  "version": "0.4.302",
+  "version": "0.4.304",
   "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 = '0.4.302';
+export const VERSION = '0.4.304';