edfcore 0.4.498 → 0.4.500
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 +33 -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 = "0.4.500";
|
|
113
113
|
//# sourceMappingURL=constants.d.ts.map
|
package/dist/constants.js
CHANGED
package/docs/CHANGELOG.md
CHANGED
|
@@ -6,6 +6,39 @@ 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
|
+
## 0.4.500
|
|
10
|
+
|
|
11
|
+
- **Added** the halves of `discontinuous.md` its existing test did not reach: what `openEdf` alone
|
|
12
|
+
reports, and what an overlap looks like.
|
|
13
|
+
- The overlap section is the one worth pinning. EDF+D never lets a record start before the previous
|
|
14
|
+
one ends; files do it anyway, and edfcore's answer is that there is no separate shape for it —
|
|
15
|
+
`EdfGap.durationSeconds` simply goes negative. The page demonstrates that with one line, `[-1, 1]`,
|
|
16
|
+
and a reader summing durations to get "time lost to gaps" needs it to be true.
|
|
17
|
+
- So the sum is asserted too: on that file it comes to zero seconds lost across a recording that
|
|
18
|
+
plainly has a hole in it, which is the warning the page prints under the example, made
|
|
19
|
+
executable.
|
|
20
|
+
- `gaps.length === segments.length - 1` is checked on the overlapping file specifically, since
|
|
21
|
+
that is the case where a reader would doubt a rule stated for "any file that has records".
|
|
22
|
+
- And the probed-index block, which is the page's other promise: nothing on the object `openEdf`
|
|
23
|
+
returns reads as "this recording is continuous" when nothing has looked.
|
|
24
|
+
|
|
25
|
+
## 0.4.499
|
|
26
|
+
|
|
27
|
+
- **Added** the scaling contract as `concepts.md` states it: `signal.scale` is `undefined`,
|
|
28
|
+
`toPhysical` throws, and `decodeDigital` still works. Three lines that are the package's central
|
|
29
|
+
refusal rather than an illustration of it — the page names what the reference implementation does
|
|
30
|
+
instead, which is to substitute a gain of 1 and return ADC counts labelled as microvolts.
|
|
31
|
+
- `physical-values.md` has its own arithmetic test and `scale.ts` has unit tests for each
|
|
32
|
+
condition. What nothing checked is the trio together on one signal, which is what a reader takes
|
|
33
|
+
away from the page: the samples are real and keep working, and it is the interpretation that is
|
|
34
|
+
unavailable.
|
|
35
|
+
- The four conditions the paragraph lists are checked as a closed list, because a fifth appearing
|
|
36
|
+
without the sentence changing is the drift worth catching, and that sentence is what a reader
|
|
37
|
+
plans around.
|
|
38
|
+
- The diagnostics block is the other closed list on that page — five properties, one per producer.
|
|
39
|
+
All five are now exercised and compared against the list parsed out of the page, so a sixth
|
|
40
|
+
producer, or a renamed property, fails here.
|
|
41
|
+
|
|
9
42
|
## 0.4.498
|
|
10
43
|
|
|
11
44
|
- **Fixed** an error message quoted on two documentation pages that no version of edfcore has been
|
package/package.json
CHANGED
package/src/constants.ts
CHANGED