edfcore 0.4.356 → 0.4.358
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/dist/constants.d.ts +1 -1
- package/dist/constants.js +1 -1
- package/docs/CHANGELOG.md +31 -0
- package/package.json +1 -1
- package/src/constants.ts +1 -1
package/dist/constants.d.ts
CHANGED
|
@@ -109,5 +109,5 @@ export declare const SIGNAL_FIELD_BLOCK_OFFSETS: {
|
|
|
109
109
|
readonly reserved: 224;
|
|
110
110
|
};
|
|
111
111
|
/** Published package version. Kept in sync with package.json by a test. */
|
|
112
|
-
export declare const VERSION = "0.4.
|
|
112
|
+
export declare const VERSION = "0.4.358";
|
|
113
113
|
//# sourceMappingURL=constants.d.ts.map
|
package/dist/constants.js
CHANGED
package/docs/CHANGELOG.md
CHANGED
|
@@ -6,6 +6,37 @@ alone does not tell you whether you were affected.
|
|
|
6
6
|
edfcore is pre-1.0. Patch releases have carried behaviour changes where the old behaviour was a
|
|
7
7
|
defect; those are called out below.
|
|
8
8
|
|
|
9
|
+
## 0.4.358
|
|
10
|
+
|
|
11
|
+
- **Added** checks for three small pure functions on `api-primitives.md`. `decodeHeaderLatin1` is
|
|
12
|
+
the one with teeth: the page does not merely say `TextDecoder` is unused, it says why, with a
|
|
13
|
+
measurement — every relevant label reports `windows-1252` on Node and decodes `0x80` as U+0080,
|
|
14
|
+
while the WHATWG standard mandates U+20AC, so the same header bytes become different strings on a
|
|
15
|
+
server and in a tab. `text-decoder-ban.test.ts` forbids the call; this checks the behaviour the
|
|
16
|
+
ban buys, at the byte the two disagree about and across all 256 of them.
|
|
17
|
+
- `formatStartTimeNaive` returns `undefined` for two conditions, and the second is the one worth
|
|
18
|
+
having: a starttime of `23.59.60` fails its grammar, and without that branch the file came back
|
|
19
|
+
as midnight — an instant it never gave, and for a sleep study the most believable start there is.
|
|
20
|
+
- Its example renders `1951`, which is the FORMAT rather than the result of the field in the
|
|
21
|
+
fixture: the two-digit year rule puts 00..84 in 2000..2084, so `02.08.51` is 2051 and only a
|
|
22
|
+
four-digit EDF+ `Startdate` can say 1951. The check compares the shape against the page and the
|
|
23
|
+
value against the rule, which is what the first version of it got wrong.
|
|
24
|
+
|
|
25
|
+
## 0.4.357
|
|
26
|
+
|
|
27
|
+
- **Added** an execution of the two-signal `trimToWindow` example on `api-primitives.md`. It exists
|
|
28
|
+
to make one point — a 256 Hz channel and a 3 Hz channel asked for the same window start at
|
|
29
|
+
different instants and hold different counts — and that is why `startSeconds` lives on
|
|
30
|
+
`EdfChunkSignal` rather than only on the chunk. Both printed rows are now produced.
|
|
31
|
+
- The rule underneath is also pinned: membership is decided against the tick edfcore PUBLISHES for
|
|
32
|
+
a sample, not against its exact rational start. The page's own worked figure is checked both
|
|
33
|
+
ways — 256 samples in a one-second record put sample 1 at 39,062.5 ticks, which is not a whole
|
|
34
|
+
tick, published as 39,063 — because selecting on the exact start excluded the very sample a
|
|
35
|
+
caller had aligned the window to, a defect this project shipped and fixed in 0.3.56.
|
|
36
|
+
- Plus the three things trimming promises about its result: a subarray view rather than a copy so
|
|
37
|
+
it allocates nothing, a zero-length result rather than an error for a window that misses, and a
|
|
38
|
+
refusal when the header is not the one the chunk was read with.
|
|
39
|
+
|
|
9
40
|
## 0.4.356
|
|
10
41
|
|
|
11
42
|
- **Added** an execution of the `resolveTimeWindow` example on `api-primitives.md`: a one-second
|
package/package.json
CHANGED
package/src/constants.ts
CHANGED