chore(papi): commit the sources behind the four polkadot-api packages (#15)

* chore(papi): commit the sources behind the four polkadot-api packages

papi-utils, bizinikiwi-bindings, metadata-builders and merkleize-metadata
held only dist/ builds, and no repository held their source. Every
source map in dist/ embeds the TypeScript it was compiled from; that is
now committed as src/ (59 files), and it matches upstream polkadot-api
(MIT) exactly once @polkadot-api/* imports are renamed:

  papi-utils           @polkadot-api/utils 0.2.0
  bizinikiwi-bindings  @polkadot-api/substrate-bindings 0.16.6
  metadata-builders    @polkadot-api/metadata-builders 0.13.0-0.13.2
  merkleize-metadata   @polkadot-api/merkleize-metadata 1.1.21-1.1.23

scripts/check-papi-sources.mjs, run in the lint job, fails unless the
source embedded in dist/ is exactly the committed src/. tsconfig.json
excludes these src/ trees; ESLint already skipped the packages. Each
README now states the upstream package and version.

* chore(papi): keep the upstream sources in upstream-src/

The pezkuwi-dev build treats any package with a src/ directory as its
own: it requires our license header on every file and builds it. These
are upstream files that must stay byte-identical to the source embedded
in dist/, so they live in upstream-src/ instead, which neither the build
nor the root tsconfig picks up.

packages/papi-utils/upstream-src/mapObject.ts

@@ -0,0 +1,29 @@
+export function mapObject<K extends string | number | symbol, I, O>(
+  input: Record<K, I>,
+  mapper: (i: I, k: K) => O,
+): Record<K, O>
+
+export function mapObject<K extends string | number | symbol, I, O>(
+  input: Record<K, I>,
+  mapper: (i: I, k?: K) => O,
+): Record<K, O> {
+  return Object.fromEntries(
+    Object.entries(input).map(
+      ([key, value]: any) => [key, mapper(value, key)] as const,
+    ),
+  ) as any
+}
+
+export type StringRecord<T> = {
+  [Sym: symbol]: never
+  [Num: number]: never
+  [Str: string]: T
+}
+
+export const mapStringRecord = <I, O>(
+  input: StringRecord<I>,
+  mapper: (value: I, key: string) => O,
+): StringRecord<O> =>
+  Object.fromEntries(
+    Object.entries(input).map(([key, value]) => [key, mapper(value, key)]),
+  ) as StringRecord<O>