edfcore 0.4.448 → 0.4.450
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 +39 -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 = "0.4.450";
|
|
113
113
|
//# sourceMappingURL=constants.d.ts.map
|
package/dist/constants.js
CHANGED
package/docs/CHANGELOG.md
CHANGED
|
@@ -6,6 +6,45 @@ 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
|
+
## 0.4.450
|
|
10
|
+
|
|
11
|
+
- **Added** a property test that `printable` never changes the width of what it is given. Every
|
|
12
|
+
table the package prints depends on it: `formatHeader` pads a label into a column, the CLI's
|
|
13
|
+
`signals` output is tab-separated for `cut -f2`, and the validation report lines observed ranges
|
|
14
|
+
up under a heading — all on text read out of a file, where EDF says a label is sixteen bytes and
|
|
15
|
+
says nothing about what may be in them.
|
|
16
|
+
- The module's docblock states the rule and rejects two alternatives, both of which are the natural
|
|
17
|
+
thing to reach for and both correct-looking: a stripped control character is gone, an escaped one
|
|
18
|
+
is legible. Either shifts a column by a character per occurrence. Those alternatives are now an
|
|
19
|
+
executable contrast rather than a comment — the property computes both and shows one shortens and
|
|
20
|
+
one lengthens.
|
|
21
|
+
- The generated text covers what a fixture-writer does not think of: astral characters, which are
|
|
22
|
+
two UTF-16 units and one iteration of a `for…of`; lone surrogates, which are one of each; the
|
|
23
|
+
Latin-1 supplement, which a European electrode label really contains; and the line separator the
|
|
24
|
+
docblock argues should pass through. DEL is drawn separately, because a generator covering only
|
|
25
|
+
0x00-0x1f leaves the second half of the condition untouched while looking thorough.
|
|
26
|
+
- Three invariants, and the third is what makes the first two mean anything: the length is
|
|
27
|
+
unchanged, no C0 or DEL survives, and every character that was not one is exactly what it was.
|
|
28
|
+
Without the third, returning a string of dots would satisfy the other two.
|
|
29
|
+
|
|
30
|
+
## 0.4.449
|
|
31
|
+
|
|
32
|
+
- **Added** a check that the sourcemaps in the published package point at files the package ships.
|
|
33
|
+
`sourceMap` and `declarationMap` in the build config and `src` in `files` are one feature: tsc
|
|
34
|
+
emits maps whose sources are `../src/x.ts` relative to `dist/` and inlines no `sourcesContent`,
|
|
35
|
+
so the maps resolve only because the TypeScript ships beside them. That is why
|
|
36
|
+
`check-tarball.mjs` refuses a tarball without `src/` with the words "the sourcemaps resolve to
|
|
37
|
+
nothing".
|
|
38
|
+
- Nothing checked that the maps say what the arrangement assumes, and three edits break it while
|
|
39
|
+
failing nothing: dropping `src` from `files` to make the tarball smaller — 250 KB of TypeScript
|
|
40
|
+
for a package whose `dist` is what runs looks like dead weight; turning `sourcesContent` on,
|
|
41
|
+
which makes `src/` redundant right up until someone turns it off again; or turning either map off.
|
|
42
|
+
- What breaks is not a build. It is a consumer stepping into `openEdf` in a debugger and landing in
|
|
43
|
+
compiled output, or "go to definition" on `EdfHeader` opening a `.d.ts` instead of the file whose
|
|
44
|
+
comments explain the field. The package still installs, still imports, still passes everything
|
|
45
|
+
here — what stops working is the reason `removeComments: false` and `declarationMap` are set at
|
|
46
|
+
all, that the source is what a reader reaches.
|
|
47
|
+
|
|
9
48
|
## 0.4.448
|
|
10
49
|
|
|
11
50
|
- **Added** tests for what `edfcore json` puts in a file you are about to pipe somewhere. Piping is
|
package/package.json
CHANGED
package/src/constants.ts
CHANGED