edfcore 0.3.64 → 0.3.65
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/CHANGELOG.md +24 -0
- package/dist/constants.d.ts +1 -1
- package/dist/constants.js +1 -1
- package/package.json +1 -1
- package/src/constants.ts +1 -1
package/CHANGELOG.md
CHANGED
|
@@ -6,6 +6,30 @@ 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
|
+
## 0.3.65
|
|
10
|
+
|
|
11
|
+
- **Fixed** `api-errors.md` publishing the wrong signature for `isEdfError` and calling a cast
|
|
12
|
+
mandatory that the compiler does not require.
|
|
13
|
+
- The page said `function isEdfError(value: unknown): value is EdfError`. It returns
|
|
14
|
+
`value is AnyEdfError` — the discriminated union over the seven concrete classes.
|
|
15
|
+
- Around that it built a **Note** explaining that "the cast in each branch is load-bearing" and
|
|
16
|
+
that "reaching for `error.budgetBytes` without the cast is a compile error", with three `as`
|
|
17
|
+
casts in the snippet to match. None of it is true: the snippet compiles without them under
|
|
18
|
+
edfcore's own `tsconfig.json`. `src/errors.ts` says why in the docblock on `AnyEdfError` — the
|
|
19
|
+
union exists precisely so that switching on `edfErrorKind` reaches the extra fields "without
|
|
20
|
+
one". The page documented the problem the union was added to solve as though it were still
|
|
21
|
+
there.
|
|
22
|
+
- The snippet is now in `tests/types/documented-examples.test-d.ts` alongside the other two, so
|
|
23
|
+
`npm run typecheck` compiles it: if a cast ever became necessary, the build would say so rather
|
|
24
|
+
than a paragraph.
|
|
25
|
+
- Adding it immediately caught a second error in the same snippet. `EdfRangeError.available` is a
|
|
26
|
+
`RecordRange`, not a count, so the helper the page calls `clampToFile(error.available)` cannot
|
|
27
|
+
take a `number`.
|
|
28
|
+
- The comparison against the page is now indentation-insensitive, because a page-level fragment sits
|
|
29
|
+
at column 0 and a compiled copy has to live inside a function. The snippet's `switch` arms were
|
|
30
|
+
reflowed to the form this project's formatter produces, so the page shows code that would survive
|
|
31
|
+
`biome check`.
|
|
32
|
+
|
|
9
33
|
## 0.3.64
|
|
10
34
|
|
|
11
35
|
- **Fixed** the pre-0.2 status text still on the website, which contradicted the package and, in one
|
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 = "0.3.
|
|
112
|
+
export declare const VERSION = "0.3.65";
|
|
113
113
|
//# sourceMappingURL=constants.d.ts.map
|
package/dist/constants.js
CHANGED
package/package.json
CHANGED
package/src/constants.ts
CHANGED