edfcore 0.5.54 → 0.5.56

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.
@@ -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 = "0.5.54";
112
+ export declare const VERSION = "0.5.56";
113
113
  //# sourceMappingURL=constants.d.ts.map
package/dist/constants.js CHANGED
@@ -79,5 +79,5 @@ export const SIGNAL_FIELD_BLOCK_OFFSETS = {
79
79
  reserved: 224,
80
80
  };
81
81
  /** Published package version. Kept in sync with package.json by a test. */
82
- export const VERSION = '0.5.54';
82
+ export const VERSION = '0.5.56';
83
83
  //# sourceMappingURL=constants.js.map
package/docs/CHANGELOG.md CHANGED
@@ -6,6 +6,46 @@ 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
+ ## 0.5.56
10
+
11
+ - **Added** the rule `precededByGap` carries, checked however the chunk was asked for. It is stated
12
+ in three places and was tested in one. `chunks.ts`: the field is `undefined` "in two different
13
+ situations: no gap, and nobody looked". `stream.ts`: a streamed chunk carries "the same
14
+ `precededByGap` a `readWindow` chunk would". `biosemi.ts` states the sharpest form — the gap
15
+ "precedes the RUN, not whichever sample the window happened to admit first, which could be a whole
16
+ record later", a rule narrowed in 0.3.67 after 0.3.92 wrote down a wider one.
17
+ - Four entry points produce chunks, and the field is computed from the index rather than from the
18
+ read — so what it says must be the same whichever asked, for the same records.
19
+ - One file with a twenty-second hole after record 4, and every way of naming the records around it:
20
+ `readRecords` by number, `readWindow` by time, `streamRecords` in pieces smaller than the run, and
21
+ `readEnvelope`. The gap appears on exactly one chunk in each — the one whose first record is
22
+ record 5 — and on none of the others, including the chunk starting at record 6, which is inside
23
+ the run and after it, and the chunk starting at record 4, which spans the hole by record number
24
+ and is not preceded by it. At four different `chunkRecords`, exactly one streamed chunk carries it
25
+ every time.
26
+ - The other of the two situations is checked beside it: on a probed index the same read gives
27
+ `undefined`, which is "nobody looked" rather than "no gap" — and the samples are identical either
28
+ way, so only the field about structure differs.
29
+
30
+ ## 0.5.55
31
+
32
+ - **Added** the window that begins before the recording does. "A window that starts before record 0
33
+ is how a pre-stimulus epoch is spelled" is the sentence `tal/ticks.ts` uses to justify `floorDiv`
34
+ and `ceilDiv` existing at all, and `annotations.md` builds a section on negative onsets around the
35
+ same idea. 0.5.15 tested the two divisions; what a caller actually writes —
36
+ `readWindow(recording, { startSeconds: -2, durationSeconds: 4 })` — was tested for its bounds and
37
+ not for the negative side of them.
38
+ - There are three outcomes and only one is an error, and it is not the one a reader expects. A
39
+ window straddling t = 0 is read, clamped to the samples that exist. A window entirely before the
40
+ recording comes back as `[]` — the same empty array a window past the end gives, which
41
+ `api-reading.md` says in as many words. Nothing throws for either: a negative bound is a
42
+ legitimate question with a possibly empty answer, and only a non-finite one is a mistake.
43
+ - The trimmed result is where it matters, and it is asserted sample for sample.
44
+ `trimToWindow(header, series, -2, 4)` starts at 0 rather than at -2 and holds exactly the samples
45
+ in `[0, 2)` — the arithmetic that needs a floor rather than a truncation, on operands that are
46
+ negative for the whole first half of the window. A bound that is not a whole second is checked
47
+ with it, and `sampleAt` is checked either side of the boundary.
48
+
9
49
  ## 0.5.54
10
50
 
11
51
  - **Added** the claim the `index` option on `validateRecording` is only worth having if it holds: a
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "edfcore",
3
- "version": "0.5.54",
3
+ "version": "0.5.56",
4
4
  "description": "Modern, typed, zero-dependency reader for EDF, EDF+, BDF and BDF+ biosignal files. Works in browsers and Node with true random access.",
5
5
  "keywords": [
6
6
  "edf",
package/src/constants.ts CHANGED
@@ -93,4 +93,4 @@ export const SIGNAL_FIELD_BLOCK_OFFSETS = {
93
93
  } as const;
94
94
 
95
95
  /** Published package version. Kept in sync with package.json by a test. */
96
- export const VERSION = '0.5.54';
96
+ export const VERSION = '0.5.56';